A consistency criterion for EQ-5D-5L bolt-on valuation

OBJECTIVE: EuroQol is developing the EQ-5D Bolt-on Toolbox, an experimental set of bolt-ons designed to improve the sensitivity of EQ-5D-5L in specific populations. For use in cost-utility analyses, EQ-5D-5L+bolt-on value sets should be consistent with EQ-5D-5L value sets. This study proposes a method to define and maintain such consistency and examines its properties. METHODS: We developed a novel contraction method that defines consistency using the distribution of disutilities in measurement data and estimates a contraction parameter from the conditional distribution of bolt-on levels given the core EQ-5D-5L profile to account for information overlap. The method was tested for nine bolt-ons: breathing, cognition, hearing, self-confidence, skin irritation, sleep, social relationships, tiredness, and vision. We used general-population samples from 14 countries, with 10,395-42,206 observations per bolt-on, and patient samples from six disease areas, with 312-2,486 observations. Core-dimension disutilities were based on the UK EQ-5D-5L value set, and bolt-on disutilities were taken from published studies. Associations of the contraction parameter with bolt-on disutilities and core-dimension correlations were assessed using linear regression. RESULTS: Achieving consistency required contraction factors ranging from 0.871 for breathing to 0.989 for sleep. Stronger contraction was required for bolt-ons with larger disutilities and for those more strongly correlated with the core dimensions, although only the former association was statistically significant. CONCLUSION: The contraction method aligns EQ-5D-5L+bolt-on value sets with EQ-5D-5L value sets, reducing potential double counting while preserving the increased sensitivity provided by bolt-ons.