Abstract
Objectives
Country-specific value sets for the EQ-5D are available which reflect preferences for health states elicited from the general population. This allows the transformation of responses on EQ-5D to health state utility values. Only twelve European countries possess country-specific value sets and no value set reflecting the preferences of Europe exists. We aim to estimate a ‘pan-European’ value set for the EQ-5D-3L, reflecting the preferences for health states of the European population that could help to evaluate health care from the perspective of the European decision-maker.
Methods
We systematically assessed and compared the methodologies of available EQ-5D-3L time trade-off (TTO) value sets from twelve European countries: Denmark, France, Germany, Hungary, Italy, Netherlands, Poland, Portugal, Romania, Slovenia, Spain and UK. Using their published coefficients, a dataset with utility values for all 243 health states was simulated. Different modelling techniques and model specifications including interaction terms were tested. Model selection was based on goodness-of-fit criteria. We also explored results with application of population size weights.
Results
Methodological, procedural and analytical characteristics of the included EQ-5D-3L valuation studies were quite comparable. An OLS based model was the preferred model to represent European preferences. Weighting with population size made little difference.
Conclusions
EQ-5D-3L valuation studies were considered of sufficient comparability to form the basis for a new ‘pan-European’ value set. The method used allows for an easy update when new national value sets become available.
Country-specific value sets for the EQ-5D are available which reflect preferences for health states elicited from the general population. This allows the transformation of responses on EQ-5D to health state utility values. Only twelve European countries possess country-specific value sets and no value set reflecting the preferences of Europe exists. We aim to estimate a ‘pan-European’ value set for the EQ-5D-3L, reflecting the preferences for health states of the European population that could help to evaluate health care from the perspective of the European decision-maker.
Methods
We systematically assessed and compared the methodologies of available EQ-5D-3L time trade-off (TTO) value sets from twelve European countries: Denmark, France, Germany, Hungary, Italy, Netherlands, Poland, Portugal, Romania, Slovenia, Spain and UK. Using their published coefficients, a dataset with utility values for all 243 health states was simulated. Different modelling techniques and model specifications including interaction terms were tested. Model selection was based on goodness-of-fit criteria. We also explored results with application of population size weights.
Results
Methodological, procedural and analytical characteristics of the included EQ-5D-3L valuation studies were quite comparable. An OLS based model was the preferred model to represent European preferences. Weighting with population size made little difference.
Conclusions
EQ-5D-3L valuation studies were considered of sufficient comparability to form the basis for a new ‘pan-European’ value set. The method used allows for an easy update when new national value sets become available.
| Original language | English |
|---|---|
| Article number | 13 |
| Journal | BMC Medical Research Methodology |
| Volume | 23 |
| Issue number | 1 |
| DOIs | |
| Publication status | Published - 12 Jan 2023 |
Bibliographical note
Funding Information:On behalf of the PECUNIA (ProgrammE in Costing, resource use measurement and outcome valuation for Use in multi-sectoral National and International health economic evaluations) project. This project has received funding from the European Union´s Horizon 2020 research and innovation programme under grant agreement No 779292.
Publisher Copyright:
© 2023, The Author(s).