Objective
We aimed to develop an EQ-5D-5L value set for Romania.
Methods
In line with the EuroQoL standardized valuation protocol, computer-assisted interviews were conducted face-to-face in a representative sample in Romania (November 2018–November 2019).
Valuation methods included composite time trade-off and discrete choice experiment tasks. Several models were tested, including models that accounted for data censoring, panel structure of the data, heteroscedasticity, conditional logit, and hybrid models. The final model was selected based on logical consistency, theoretical considerations, and use of all available data. We compared our value set with other value sets from Central and Eastern Europe region.
Results
Data from 1493 respondents was used to estimate the value set. A censored hybrid model corrected for heteroscedasticity was selected to represent the value set. The highest decrements in utility were observed for the pain/discomfort dimension (0.375), followed by the mobility dimension (0.293). Health utilities ranged from 1.000 to − 0.323 and 1.3% of the values were negative. The model was corrected with survey weights to better reflect the representativeness of the sample, but the first two coefficients of the self-care dimension stopped being logically consistent. Differences were found between the Romanian, Hungarian and Polish EQ-5D-5L value sets. Good agreement was noted with the Romanian EQ-5D-3L value set, with a swap between pain/discomfort and mobility in ranking of dimensions.
Conclusion
A value set for EQ-5D-5L is now available for Romania. This will push one-step further the development of health technology assessment and encourage more health-related quality-of-life research to be conducted locally.