Presentation (CTI)

OBJECTIVES: Reporting “no problems” across all five EQ-5D-5L dimensions is referred to as “full health”. EQ-5D-5L full-health prevalence could vary substantially across countries. Using the multinational EuroQol EQ-DAPHNIE dataset, we explored how country, sociodemographic, health, behavioural, and healthcare characteristics were associated with full-health reporting.
METHODS: The EQ-DAPHNIE project collected general-population data from 15 countries across Europe, Asia, North and South America, and Oceania, using standardized survey methods. The survey included EQ-5D-5L alongside four variable sections: sociodemographic characteristics, health (e.g., comorbidities), behaviours (e.g., physical activity), and healthcare variables (e.g., outpatient use). We assessed cross-country consistency of crude full-health prevalence gradients across these variables by comparing the direction and rank ordering of categories. We then estimated a probit model with full health as the outcome. Predictors were selected from the four variable sections, with country as a fixed variable.
RESULTS: Of 68,416 respondents, 67,178 (98%) had complete EQ-5D-5L data; 21,289 (31.7%) reported full health. Crude prevalence ranged from 20.3% (Canada) and 23.2% (U.S.) to 46.0% (China) and 47.5% (Japan). Full-health prevalence gradients for health, behavioural, and healthcare variables were broadly consistent across countries, whereas only 5 of 16 sociodemographic variables showed a consistent direction (e.g., income status). In the probit model (n=52,717; McFadden R²=0.3), partial adjustment for 15 health variables attenuated Japan's and China's full-health probability by 7-8 percentage points (pp) each, while Canada and the U.S. increased by 6 pp. Among sociodemographic, behavioural, and healthcare variables, social network support showed the largest independent association, spanning 12 pp in full-health probability between the highest and lowest support levels. After full adjustment, predicted full-health probabilities ranged from 26.2% (Canada) to 42.7% (Japan).
CONCLUSIONS: Reporting EQ-5D-5L full health is not fully comparable across countries. Substantial country-level differences remained after adjustment for health, personal, and contextual factors. International applications of the EQ-5D-5L should account for individual- and country-level variation.