Health utility values associated with vision impairment among the elderly in Hong Kong : a community-based study using EQ-5D-5 L and time trade-off

Abstract

Background: To estimate health utility decrements associated with correctable and uncorrectable vision impairment (VI) in community-dwelling older adults in Hong Kong using the EQ-5D-5 L and time trade-off (TTO), and to compare the sensitivity of these utility instruments in detecting VI's quality-of-life impacts. Methods: A random sample of 999 older adults, previously screened in a community-based eye care programme between 2015 and 2017 and not referred for eye specialist care, completed follow-up eye examinations and questionnaire surveys between 2022 and 2024. Distance VI was defined as presenting visual acuity worse than 6/12 in the better-seeing eye, and categorized as correctable (due to uncorrected refractive error, URE) or uncorrectable (due to pathology), based on best-corrected visual acuity. Health utility was measured using the EQ-5D-5 L questionnaire and the TTO method. Associations between VI type and utility scores were analyzed using multivariable linear regression. Results: The prevalence of VI was 19.1%, with 15.4% correctable and 3.7% uncorrectable. Mean EQ-5D-5L utility scores were 0.91 (standard deviation [SD] = 0.15) for those without VI, 0.90 (SD = 0.12) for correctable VI, and 0.85 (SD = 0.22) for uncorrectable VI (p < 0.05). Mean TTO scores followed a similar pattern: 0.95 (SD = 0.14), 0.94 (SD = 0.15), and 0.89 (SD = 0.22), respectively. In adjusted models, uncorrectable VI was significantly associated with a 0.052 decrease in TTO utility (95% confidence interval: −0.103 to −0.002; p < 0.05), while correctable VI showed no significant utility loss in either measure. Conclusions: There were generally lower utility values for community dwelling elders with VI compared to no VI, with significant reductions of uncorrectable VI due to eye diseases in health utility among older adults living in the community. In contrast, correctable VI due to URE did not significantly affect utility values, suggesting that generic instruments like EQ-5D-5 L and TTO may lack sensitivity to detect the burden of milder, reversible vision loss. These findings provide essential data for evaluating the cost-effectiveness of community eye care programs and highlight the need for more targeted quality-of-life measures in vision research.