OBJECTIVES: Curative therapies have the potential to restore patient health and significantly improve survival. However, cure assumptions in HTA appraisals can be difficult to model and support with evidence. The aim of this research was to investigate how cure assumptions have been clinically justified, modelled, and considered in recent NICE technology appraisals across all therapy areas.

METHODS: A targeted search was conducted of all NICE single technology appraisals published in the last two years which included cure assumptions. Appraisals were excluded if a cure assumption was not explicitly modelled. Data extracted included details of the cure assumption, clinical data supporting the cure assumption, cure modelling approach and HTA feedback on the cure assumption.

RESULTS: A total of 11 appraisals were identified in the search. Although the search was not limited to oncology therapies, all 11 appraisals were in oncology. Most appraisals (55%, n=6) applied a cure assumption after patients were progression-free for 5 years, while 27% (n=3) of appraisals used 2 years. The remaining two appraisals applied cure assumptions at 3 years and 30 months. There were two separate approaches to cure modelling noted. In one approach, general population mortality rate was assumed after the cure timepoint, which was applied in the majority of appraisals (73%, n=8). The other approach was the development of a mixture-cure model which was applied in 27% (n=3) of appraisals. Finally, in all 11 appraisals both the ERG/EAG and NICE considered the cure assumption uncertain, with varying degrees of acceptability.

CONCLUSIONS: All NICE appraisals in the last two years where cure assumptions have been applied were in oncology. Despite variability in data informing the assumption and modelling approach, NICE and the ERG/EAG typically considered the assumption to be highly uncertain. However, in most cases the uncertainty around the cure assumption was accepted if the medicine was cost-effective.