Scale and rate heterogeneity in the EQ-5D-5L valuation

Area of Science:

• Health Economics
• Psychometrics
• Biostatistics

Background:

• Estimating quality-adjusted life year (QALY) values often involves ancillary parameters like scale factors and discount rates.
• These parameters can vary significantly across individuals, impacting the accuracy of QALY estimations.
• Understanding and modeling this heterogeneity is crucial for robust health valuation.

Purpose of the Study:

• To investigate the impact of allowing heterogeneity in scale and discount rate parameters on the estimation of EQ-5D-5L values.
• To assess whether accounting for individual differences in these parameters improves the face validity of QALY value sets.

Main Methods:

• Utilized data from two waves of the 2016 EQ-5D-5L valuation study (N=1017 and N=1229).
• Estimated conditional logit (CL) and mixed logit models, including random scale, random rate, and bivariate models.
• Hypothesized and tested for correlation between scale and rate parameters and their impact on model validity.

Main Results:

• Scale and rate parameters demonstrated significant heterogeneity and a positive correlation.
• Allowing for heterogeneity reduced the number of insignificant incremental effects in the EQ-5D-5L value set (from 6 to 2 out of 20).
• Estimates from CL and bivariate mixed logit models were highly correlated and concordant.

Conclusions:

• Incorporating heterogeneity in scale and rate parameters enhances the face validity of EQ-5D-5L values.
• This approach adds three parameters (two variances and a correlation) to the estimation model.
• Findings have implications for experimental design and choice analysis in health valuation research.