OBJECTIVES: To review NICE (National Institute for Health and Care Excellence) HST (Highly Specialised Technology) appraisals and determine the most commonly reported uncertainties relating to health-related quality of life (HRQoL) and disease burden for patients & caregivers when modelling rare diseases.

METHODS: The company submissions for 24 HST appraisals were reviewed to investigate the methods used to capture patient & caregiver HRQoL and disease burden, and identify key areas of uncertainty. The evidence review group (ERG) and committee meeting reports were reviewed to determine common criticisms and elements influencing the committee decisions.

RESULTS: Companies seldom used the EQ-5D collected in disease-specific clinical studies (recommended by NICE) to estimate patient & caregiver HRQoL, citing biases with single-arm studies, small patient numbers and limited sensitivity of generic measures in rare diseases. Vignette studies were considered acceptable alternatives if validated by clinicians and completed by the patient. Utility values from proxy conditions were also considered appropriate if sufficient evidence was provided to display similarities between disease characteristics and symptoms. Approximately 70% of submissions reviewed quantified caregiver disutility but concerns around the methods adopted were frequently reported by ERGs and committees (e.g. the inappropriate inclusion of bereavement disutilities). The quantification of patient and caregiver burden was deemed acceptable if presented as separate scenario analysis. However, companies were often criticised for overestimating productivity loss when using the human capital approach, rather than the friction cost method.

CONCLUSIONS: All technologies from fully completed submissions were recommended in some capacity by NICE (e.g. accompanying managed access agreement) thereby indicating that committees are understanding of the substantial uncertainties associated with modelling patient and caregiver HRQoL and disease burden in rare diseases. Nevertheless, common themes were identified that may influence the approach adopted for future appraisals (e.g. appropriate methodology for vignette studies and proxy utilities).