OBJECTIVES: Replacement of the NICE end-of-life criterion decision modifier with a tiered disease severity criterion based on proportional/absolute QALY shortfall is a significant change in willingness-to-pay policy. Of 19 eligible NICE technology appraisals only 2 have achieved a QALY weighting (1 x 1.2 and 1 x 1.7) largely by deterministic analysis. We evaluated the extent to which NICE’s deterministic approach to disease severity classification could be impacted by parameter uncertainty.

METHODS: A review screened 1294 abstracts for published between 2013 and 2019 of which 337 UK-based analyses with lifetime horizons were analysed. Among these, 20 studies (1.5%) reported quasi-lifetime QALYs with uncertainty (standard error [SE] or 95% confidence interval [95%CI]). Deterministic QALY shortfall was assessed using the UK population norms for life-expectancy and age-specific utility recommended by NICE, while probabilistic QALY shortfall was calculated for each weighting threshold (x1.0, x1.2, and x1.7) by using the SE/95%CI to estimate the probability that diseased cohorts would meet the criteria for a proportional or absolute QALY shortfall.

RESULTS: Forty diseased cohorts representing usual care in 27 distinct diseases were identified from the selected studies with baseline ages ranging from newborn to 83 years. Most (n=33) were deterministically classified with no QALY weighting (x1.0), 3 cohorts (spinal injury and advanced heart failure) merited a x1.2 weighting, and 4 (inborn errors of metabolism) achieved a x1.7 weighting. Probabilistic classification showed only 5 cohorts where the likelihood of corresponding deterministic classification was less than 90%.

CONCLUSIONS: In the available sample good alignment was observed between the deterministic and probabilistic approaches to QALY shortfall assessment. Few peer-reviewed cost-effectiveness studies using the UK reference case report lifetime QALYs with associated uncertainty, which will hamper a priori efforts to understand the likelihood of disease severity classification by NICE stakeholders.