Reliability, validity and responsiveness of the Greek MacDQoL individualized measure of the impact of macular degeneration on quality of life

Purpose To assess the psychometric properties of the Greek Macular Disease-Dependent Quality of Life Questionnaire (MacDQoL). Methods The MacDQoL was translated in Greek and administered to 191 patients with neovascular age-related macular degeneration (AMD). To assess validity, all patients completed the Greek SF-12 health survey and underwent vision measurements. For test-retest reliability, a subset of twenty participants completed the MacDQoL twice, 2 weeks apart. Responsiveness was assessed on 102 patients who completed the MacDQoL at a follow-up visit, 1 year later. Rasch analysis was used to assess the Greek MacDQoL's response category functioning, precision, unidimensionality, targeting and differential item functioning. Results Internal and test-retest reliability of the average weighted impact (AWI) was 0.952 and 0.97, respectively. Test-retest reliability of MacDQoL items ranged from 0.78 to 0.99. Principal component analysis revealed three subscales (activities, embarrassment and family life), which were also confirmed by confirmatory factor analysis. Rasch analysis revealed poor functionality of response categories and that was resolved by collapsing response categories and using the impairment scores only. In terms of convergent validity, the AWI and revised MacDQoL scales showed significant correlations with SF-12 summary scales (ρ = 0.21-0.30) and vision assessments (ρ = 0.31-0.46). Poorer AMD-related QoL at 1-year follow-up was associated with deterioration in distance visual acuity and worse eye near visual acuity. Conclusions The Greek MacDQoL is a reliable, valid and sensitive to change in vision instrument for assessing AMD patients' perceptions of QoL. However, Rasch analysis revealed that its multiplicative rating scale is flawed. Scientific measurement was restored with a number of revisions to the questionnaire.