OBJECTIVES

Focusing on the National Institute for Health and Care Excellence (NICE)’s short life expectancy end-of-life (EoL) criterion, we aimed to examine the committee’s preferred statistic for estimating life expectancy, while assessing how the innovative nature of the appraised technology influenced the committee’s decision.

METHODS

Submissions published in 2018/19 where manufacturers have applied for EoL were closely reviewed. The rationale for using the mean or median to estimate the life expectancy, along with the innovative nature of the drug and the level of support from the committee, were extracted.

RESULTS

Of the 16 identified submissions, 10 satisfied the EoL criteria; eight were recommended via the Cancer Drugs Fund and two via routine commissioning. Overall, the committee preferred to estimate life expectancy using the mean from extrapolated overall survival (OS), with reported median(s) used supportively. However, in TA567, with a considerable disparity between the mean and median estimates (43 vs 4 months) and across curve fits, as well as precedent for EoL in another submission, the committee was comfortable in making an exception with the drug being innovative and aimed for a population of patients with high unmet needs. Of the six submissions where the short life expectancy condition was not met, none were viewed as highly innovative and in each case the modelled mean OS exceeding 24 months (range, 30 – 44 months) was given as the reason the EoL criteria were not met.

CONCLUSIONS :

When estimating life expectancy, we recommended that, in line with the economic analysis, the extrapolated mean OS is used as the primary evidence, with the reported median from the trial(s) or other relevant published studies used supportively. Where the mean OS exceeds 24 months, EoL status is unlikely to be granted unless there is precedent in the same disease area and the drug is highly innovative.