The estimation of health state utility values in rare diseases: do the approaches in submissions for NICE technology appraisals reflect the existing literature? A scoping review

Background: Rare diseases negatively impact patients' quality of life, but the estimation of health state utility values (HSUVs) in research studies and cost-utility models for health technology assessment is challenging.

Objectives: This study compared the methods for estimating the HSUVs included in manufacturers' submissions of orphan drugs to the National Institute for Health and Care Excellence (NICE) with those of published studies addressing the same rare diseases to understand whether manufacturers fully exploited the existing literature in developing their economic models.

Methods: All NICE Technology Appraisal (TA) and Highly Specialized Technologies (HST) guidance documents of non-cancer European Medicines Agency (EMA) orphan medicinal products were reviewed and compared with any published primary studies, retrieved via PubMed until November 2020, and estimating HSUVs for the same conditions addressed in manufacturers' submissions.

Results: We identified 22 NICE TA/HST appraisal reports addressing 19 different rare diseases. Sixteen reports presented original HSUVs estimated using EQ-5D or Health Utility Index (n = 12), direct methods (n = 2) or mapping (n = 2), while the other six included values obtained from the literature only. In parallel, we identified 111 published studies: 86.6% used preference-based measures (mainly EQ-5D, 60.7%), 12.5% direct techniques, and 2.7% mapping. The collection of values from non-patient populations (using 'vignettes') was more frequent in manufacturers' submissions than in the literature (22.7% vs. 8.0%).

Conclusions: The agreement on methodological choices between manufacturers' submissions and published literature was only partial. More efforts should be made by manufacturers to accurately reflect the academic literature and its methodological recommendations in orphan drugs submissions.