Reliability and validity of the Turkish EQ-5D-5L in adults living with HIV

The advent of antiretroviral therapy (ART) has shifted human immunodeficiency virus (HIV) care from survival toward optimizing quality of life. In the contemporary ART era, health-related quality of life (HRQoL) is recognized as a core outcome in HIV care. The EQ-5D-5L is widely used to measure HRQoL. However, its measurement properties have not been evaluated in Turkish people living with HIV (PLHIV). This study aimed to evaluate the reliability and validity of the Turkish EQ-5D-5L in PLHIV.

We conducted a psychometric validation study at a tertiary center and enrolled 157 PLHIV. For test-retest reliability, 60 participants completed the EQ-5D-5L on two occasions. Because Turkey currently lacks an EQ-5D-5L value set, and to enhance the robustness of interpretation , utility scores were calculated using five international value sets. We assessed internal consistency, test-retest reliability, internal structure using confirmatory factor analysis (CFA), and construct validity using known-groups comparisons and within-instrument convergent and discriminant validity analyses.

The cohort had a median age of 41 years, was predominantly male (93%), mostly virally suppressed (94.3%), and had generally favorable immunological status. A ceiling effect was observed in 36.9% of participants. Internal consistency was high (ordinal α = 0.930; ω total = 0.954). Test-retest reliability was good for utility indices (intraclass correlation coefficient (ICC) = 0.799-0.886) and the level sum score (ICC = 0.880), but lower for the EuroQol Visual Analogue Scale (EQ-VAS) (ICC = 0.656). Model fit was good (Comparative fit index (CFI) = 0.999; root mean square error of approximation (RMSEA) = 0.053), and standardized loadings ranged from 0.651 (for anxiety/depression) to 0.975 (for usual activities). Known-groups comparisons showed significantly lower utility scores and EQ-VAS values in participants with comorbidities (p < 0.05). Within-instrument convergent validity was supported by moderate correlations between utility indices and EQ-VAS (ρ = 0.602-0.615).

The Turkish EQ-5D-5L demonstrates acceptable reliability and construct validity in PLHIV receiving routine care, and CFA suggested a coherent internal response structure in this sample. Ceiling effects in this clinically stable cohort warrant consideration when selecting outcomes for studies targeting subtle within-person change. Development of a Turkey-specific value set remains a priority.