Psychometric properties of health-related quality of life tools for clinically stable people living with HIV in Cambodia

In Cambodia, treatment outcomes for people living with HIV are routinely monitored using clinical indicators; however, these measures do not capture the broader impact of HIV and its treatment on health-related quality of life (HRQoL). HRQoL encompasses an individual’s perceived wellbeing and functioning across multiple domains and is commonly assessed using instruments such as the EuroQol five-dimension (EQ-5D) and World Health Organization Quality of Life-BREF (WHOQOL-BREF). This study evaluated the psychometric properties of the Khmer versions of these instruments among people living with HIV in Cambodia. This study was nested within a quasi-experimental evaluation of a community-based antiretroviral therapy delivery study. The HRQoL instruments were pre-tested among 28 people living with HIV, and subsequently administered to 500 clinically stable people living with HIV by trained data collectors. Sociodemographic characteristics, comorbidities, and depressive symptoms were also collected. Psychometric evaluation included assessments of construct validity (structural, convergent, and known-groups) and internal consistency. Ceiling effects were observed for the EQ-5D-5L. Confirmatory factor analysis provided reasonable but not definitive support for the conceptual structure of WHOQOL-BREF. Convergent validity was partially supported by moderate correlations between WHOQOL-BREF and EQ-5D-5L (|ρ|=0.30–0.38), and between CES-D-10 with both instruments (|ρ|=0.36–0.42). However, theoretically unexpected correlations were also observed. On known-groups validity, EQ-5D-5L utility and EQ VAS showed lower HRQoL among participants with comorbidities; however, the adjusted differences were not statistically significant. WHOQOL-BREF findings were mixed, with higher scores observed in the physical health and social relationships domains among participants with comorbidities. Internal consistency was acceptable-to-good for WHOQOL-BREF domains overall (α = 0.71–0.80), except for the domain on social relationships (α = 0.61). This study provides preliminary evidence regarding the psychometric properties of the Khmer versions of the EQ-5D-5L and WHOQOL-BREF among clinically stable people living with HIV in Cambodia. However, the mixed findings highlight the need for further validation before these instruments can be used to inform patient-centred care and health policies for people living with HIV in Cambodia. Future studies should investigate the psychometric properties of these instruments across diverse HIV subpopulations in Cambodia, and develop a Cambodia-specific value set to better capture local societal preferences.