Objectives
This study aimed to evaluate the psychometric performance, including test-retest reliability, ceiling/floor effects, and construct validity of the interviewer-administered (IA) and self-completion (SC) versions of EQ Health and Well-being instrument (EQ-HWB) and EQ-HWB-9.
Methods
A longitudinal survey in Bandung, Indonesia, recruited 300 respondents using stratified quota sampling, comprising 200 literate individuals, 50 with low literacy or illiteracy, and 50 patients. Respondents completed EQ-HWB, EQ-5D-5L, and Warwick-Edinburgh Mental Well-being Scale using either IA or SC modes. Psychometric analyses included ceiling/floor effects, convergent and known-group validity, and test-retest reliability (Gwet’s AC2, intraclass correlation coefficients).
Results
The final sample (mean age 39.22 ± 15.81, 62.67% female) demonstrated ceiling effects on EQ-HWB items ranging from 24.0% (“Exhaustion”) to 86.33% (“Hearing”), with 5 items exceeding the 70% threshold. In contrast, EQ-HWB-9 showed lower ceiling effects, with only 1 item (“Getting around inside and outside”) reaching the ceiling criterion (71.67%). Strong IA-SC agreement was observed (eg, AC2 = 0.95 for “Getting around inside and outside”). In the illiterate subgroup, item-level agreement remained strong, and instrument-level reliability was acceptable, supporting the feasibility of interviewer administration in low-literacy populations. Convergent validity was established with EQ-5D-5L and Warwick-Edinburgh Mental Well-being Scale. Known-group comparisons based on EQ-VAS (<80 vs ≥80) and patient status showed significant discrimination.
Conclusions
To our knowledge, this is the first study to validate the experimental IA versions of EQ-HWB and EQ-HWB-9 in Indonesia. The findings support their robust psychometric properties and feasibility for use in low-literacy and patient populations.
扫码关注我们
求助内容:
应助结果提醒方式:
