OBJECTIVES: This study aims to investigate the predictive properties between a Visual Analogue Scale (VAS) and the EuroQol EQ-5D-5L utility scores in End-Stage Liver Disease (ESLD) patients across European countries as part of the EU COPE transplantation trial.

METHODS: Complete case data was extracted from the EU COPE liver transplantation trial (n=79) across three time points. Country-specific EQ-5D-5L normative values were used to calculate utility scores. Pearson’s correlation coefficient assessed relationships between utility scores and the VAS scores. Age was the predictor for two linear regression models that explored its influence on the VAS and utility scores. A pooled analysis of all time points evaluated the predictive power of the VAS on utility scores.

RESULTS: In the UK, correlations ranged from 0.55 to 0.62 (p < 0.001). Spain and Belgium showed correlations above 0.70 (p < 0.05), while Germany had weaker correlations, especially at 30 days post-transplant (p = 0.054). Age significantly affects VAS scores in the UK, Spain, and Belgium. Utility scores demonstrated significant age effects in the UK and Belgium, but not in Germany and Spain. When assessing model fit, R2 values range from 34.6% to 41.5%, which signifies moderate predictive model given limitations on available controls and confounding factors. Therefore, it suggests that utility scores explain a moderate amount of the variability in EQ-VAS overall.

CONCLUSIONS: The study found that utility and VAS scores are associated, though the correlation strength varies across countries. Age significantly affects both measures, contributing to variation, which suggests age-related stratification is necessary for ESLD quality of life modelling. The moderate and variable predictive power of utility scores may indicate that the two techniques cannot be considered as interchangeable. Further research is warranted in understanding computational approaches addressing the impact of clinical heterogeneity for assessing quality of life.