Valuing Health States: Methods and Findings from PRET and PRET-AS

This report, authored by Brendan Mulhern and colleagues, presents the methodology and findings of the “Preparatory study for the Re-valuation of the EQ-5D Tariff project” (PRET) and its extension, “PRET – Additional Sample” (PRET-AS). Published in Health Technology Assessment in February 2014, the study addresses the critical need for an updated population value set (or “tariff”) for the EQ-5D, a widely used generic measure of health-related quality of life (HRQL).

Background and Purpose: The current UK EQ-5D tariff, based on the 1994 Measurement and Valuation of Health (MVH) study, utilized face-to-face interviews and the time trade-off (TTO) method. However, significant developments over the past 15 years necessitated a revaluation:

• Recognition of MVH TTO shortcomings, particularly for states considered “worse than dead”.
• Advances in health-state valuation methods, such as Discrete Choice Experiments (DCEs) and the new DCE incorporating duration (DCETTO).
• New modes of administration, like computer-assisted personal interviews (CAPIs) and online surveys.
• Development of the five-level EQ-5D (EQ-5D-5L), offering greater descriptive detail than the original three-level version.

The PRET project, funded by the Medical Research Council (MRC)-National Institute for Health Research (NIHR) Methodology Research Programme and the EuroQol Group, aimed to explore a range of methodological issues associated with these health-state valuation techniques to inform the generation of up-to-date EQ-5D-5L population value sets.

Study Stages and Key Findings: The research was conducted in four stages surveying over 6500 respondents:

1. Stage 1 (PRET & PRET-AS online surveys): Investigated eight methodological topics using large-scale online binary choice questions. Findings indicated that health-state preferences are not independent of duration, person perspective, lead time duration, when health events take place (time preference), or satisfaction associated with the state. It also found that DCETTO is a valid method for generating EQ-5D-5L utility values, producing logically ordered coefficients and a unimodal range of predicted values without arbitrary transformations or exogenous anchoring. Binary choice LT-TTO was also found potentially feasible but required further development.
2. Stage 2 (Online and CAPI comparison): Compared online and face-to-face CAPI administrations of identical binary choice tasks. It concluded that both modes produced comparable results for these tasks, despite demographic differences in the samples.
3. Stage 3 (CAPI investigation of task acceptability): Used CAPI surveys to explore the acceptability and completion of TTO, LT-TTO, and DCETTO tasks. Respondents generally found TTO and LT-TTO easier to complete than DCETTO. This stage also revealed uncertainty among some respondents regarding distinguishing between EQ-5D-5L levels 4 (“severe”) and 5 (“extreme”), and that personal and subjective factors influence responses.
4. Stage 4 (Qualitative investigation): Employed a “think-aloud” method during interviews to gain in-depth understanding of respondents’ thought processes for binary choice and iterative health-state valuation exercises. This further highlighted respondent strategies, the influence of personal life and health experiences, and confirmed difficulties in differentiating between severe and extreme EQ-5D-5L levels.

The PRET and PRET-AS projects significantly advanced the knowledge base on conducting health-state valuation studies, particularly for the EQ-5D-5L. While the EuroQol Group’s official protocol for EQ-5D-5L valuation ultimately adopted CAPI-based, face-to-face interviews using iterative TTO and DCE without duration, the findings from PRET informed this development and highlighted the potential of innovative binary choice methods like DCETTO for future research and experimental applications. The project underscored the complexity of health-state preferences, demonstrating their dependence on various factors and the importance of careful study design and respondent engagement.

Reference: Mulhern, B., Bansback, N., Brazier, J., Buckingham, K., Cairns, J., Devlin, N., Dolan, P., Hole, A. R., Kavetsos, G., Longworth, L., Rowen, D., & Tsuchiya, A. (2014). Preparatory study for the revaluation of the EQ-5D tariff: methodology report. Health Technology Assessment, 18(12). https://doi.org/10.3310/hta18120