We assessed the HRQOL of HIV/AIDS patients attending ART clinics using the EQ-5D-5L profile, single index, and VAS, and determined the association of socioeconomic and clinical characteristics and HRQOL. The instrument showed an excellent reliability and good discriminative validity. We found a high proportion in Anxiety/Depression and Pain/Discomfort among HIV/AIDS patients who were taking ART. The health utility of patients was 0.65 and 0.70 for EQ-5D-5L single index and VAS, respectively. Gender, education, and employment were significantly associated with HRQOL. Advanced HIV/AIDS patients, poor immune status, having AUD and drug use, were associated with poorer HRQOL.
This is the first study examining psychometric properties of the new version - EQ-5D-5L in HIV/AIDS patients. It showed that the reliability was 0.85, which was close to the threshold of 0.90, indicating the potential use of EQ-5D-5L for measuring HRQOL at an individual level. In the literature, applications of the EQ-5D-3L in HIV/AIDS showed its convergent validity with the MOS-HIV, and discriminative validity with AIDS-defining events, disease severity (CD4, viral load, and HIV/AIDS stages) [35, 36, 39]. We observed the same validity of the EQ-5D-5L in discriminating patients at different CD4 groups, and HIV/AIDS stages. In addition, we found that EQ-5D-5L could distinguish patients at different clinically meaningful duration of ART. Another important characteristic of the EQ-5D-5L was that it had a smaller ceiling effect and average HRQOL score of the HIV/AIDS population than the previous 3-level version . Therefore, there will be room for assessing the improvements of HRQOL overtime if one would like to use this tool for longitudinal assessments. Consequently, findings of this study contribute to the cumulative evidence of measurement properties of the EQ-5D instruments in HIV/AIDS population that may inform the selection of measures in both cross-sectional and follow-up designs.
Compared with previous studies, we found some similar influential factors of HRQOL in HIV/AIDS patients, for example, employment was positively associated with HRQOL [27, 33]. As for the duration of ART, there was a consistent finding that patients might experience HRQOL reduction during the ART compared to those not-yet-eligible for ART . This could be due to the immune deterioration and negative impact of ART side effects. Moreover, we found that changes in HRQOL during ART were non-linear: it decreased within the first year of ART, but then increased afterwards. We also confirmed the negative impacts of alcohol use and drug use on HRQOL outcomes of ART for HIV/AIDS patients [24, 27]. In practice, both alcohol and drug use are known to be associated with delayed access to health care, suboptimal adherence, more severe co-morbidities, and poorer outcomes of ART [22, 24].
This study has helpful implications for both HRQOL assessments and ART services. The measurement properties of the EQ-5D-5L holds potentials in monitoring changes in HRQOL, which are associated with meaningful clinical indicators, such as, CD4 thresholds, HIV/AIDS progression, or responses to ART treatment. Moreover, it helps identify areas for interventions to improve the health outcomes of patients. In this study, we observed a very high prevalence of reported problems in Anxiety/Depression that suggests the necessity of psychological support during ART for HIV/AIDS patients. As for the patient management, the 1st year ART is an important period when we observed a significant reduction in HRQOL. During this period, patients had to adapt with the strict compliance, residual opportunistic infections, and side effects of antiretroviral medications. The adherence should also be maintained during more stable periods of ART afterwards. In the Vietnamese settings, stigma and discrimination were found to be a significant barrier that affected adherence to and HRQOL outcomes of ART . The reduction in HRQOL observed in women suggests that gender-specific impact mitigation and support interventions should be in place. Previous works identified that peer’s support, vocational training, job referrals, and microfinance are potential interventions to support women with HIV/AIDS [27, 33, 40–42].
The strengths of this study include the involvement of patients in three epicenters across levels of health systems and geographical areas of Vietnam. In addition, we employed the Thailand tariff for deriving the EQ-5D-5L index, which has close culture and other characteristics with Vietnam. Besides, there were some limitations that should be acknowledged. First, we did not have viral load assessments, a gold standard in monitoring ART outcomes, since viral load tests are costly and not regularly assessed for all HIV/AIDS patients in Vietnam. Second, tests of convergent validity compared only the measures derived from the same EQ-5D instrument. Moreover, the cross-sectional design might be less capable for causal inference as well as for assessing the responsiveness of the EQ-5D-5L in this measurement. In addition, patients at clinics were selected conveniently making the sample not representative for the population of HIV/AIDS patients and limiting the generalizability of study findings.
In conclusion, the EQ-5D-5L showed good properties in measuring HRQOL of HIV/AIDS patients and has potentials for monitoring ART outcomes. Integration of HRQOL measurement using EQ-5D-5L in HIV/AIDS clinical practice could be helpful for economic evaluation of HIV/AIDS interventions. Further research to develop a scoring system of the Vietnamese population is recommended.