Quality review of a proposed EQ5D-5L value set for England

Background and objectives

 EQ-5D is an instrument for measuring and valuing health related quality of life. It comprises a classification system that allows respondents to describe their health, and a set of associated “utility” values for those health states based on preferences of the general public which allow the calculation of Quality Adjusted Life Years (QALYs). The descriptive system covers five dimensions: mobility, ability to self-care, ability to undertake usual activities, pain and discomfort, and anxiety and depression.

 Until recently, EQ-5D allowed respondents to respond at one of 3 levels (3L) for each dimension (no problems, some problems, extreme problems). The 3L is the most widely used preference based measure used in economic evaluation in England, due in part to the fact that NICE recommends use of 3L health utilities in economic evaluations submitted to its Technology Appraisal Programme (NICE, 2013).

 A new version of the instrument, EQ-5D-5L, aims to improve measurement by using five levels of severity for each dimension (no problems, slight problems, moderate problems, severe problems, and extreme problems). Utilities have been published for the 5L for England (Devlin et al. 2018). It has been shown that economic evaluations undertaken using 5L rather than 3L are likely to generate very different results (Wailoo et al., 2017, Hernandez et al. 2017), so 3L and 5L cannot be used interchangeably if consistent decision making is required.

 Before a decision can be made about 3L versus 5L for economic evaluations that inform UK health policy, there is a requirement to critically assess the data and statistical methods that were used to generate the 5L value set. This reports contains this quality assurance assessment.

The Valuation Methodology

 The EuroQol Group has developed and promulgated a valuation protocol known as the EuroQol Valuation Technology (EQ-VT) for EQ-5D-5L. The EQ-VT has two main elements: a set of ten lead-time time trade-off (TTO) experiments and a set of seven discrete choice (DC) experiments, which require members of the public to compare and evaluate hypothetical health states. EQ-VT specifies the numbers of experimental subjects, the experimental tasks and health states to be valued and provides a digital environment for computer-assisted personal interviewing. The TTO and DC experiments involve only a small proportion of the 3,125 logically possible 5L health states; statistical modelling is used to extrapolate the observed data to the remaining health states.

 The English valuation assessed in this report was based on EQ-VT version 1.0. Serious concerns with data quality were subsequently identified from the five countries that used version 1.0. EQ-VT 1.0 has now been superseded and is no longer in use by EuroQol, so there is good reason to investigate the reliability of the data underlying the published value set.