Abstract
Abstract: :
Purpose: To describe the responsiveness of the NEI-VFQ to the Age-Related Eye Disease Study (AREDS) outcome variables: progression to advanced AMD, progression of lens opacity, and decrease in visual acuity(VA). Methods: The 39-item NEI-VFQ was administered and the severity of the 3 major outcome variables was assessed at 2 AREDS visits 1-4 yrs apart (median 3 yrs) for 4,142 participants. Events, in either eye, were defined as: (1) age-related macular degeneration (AMD) – development of or treatment for neovascular AMD or development of central geographic atrophy; (2) lens opacity – nuclear: 1.5 unit increase (0.9-6.1 unit scale), cortical: 10% increase in area, PSC: 5% increase in area; (3) VA loss of ≥15 letters. Responsiveness was measured by: t-statistic, effect size (ES), responsiveness statistic (SRM) and area under the receiver operating characteristic curve (AUC). A variance component model was used to estimate each ocular factor's contribution to the overall variability of the NEI-VFQ score. Results: Of the 4,142 participants with 2 VFQ administrations, 3,687 had fundus photographs available at each administration, 3,396 had at least one natural lens present and lens photographs at each administration, and 3,402 had VA of 20/200 or better at the first administration. Numbers of events of each type are: 342 AMD, 492 lens opacity, and 409 VA loss. The overall NEI-VFQ score was responsive for AMD progression (t=14.7, P<0.001; ES=0.86; SRM=1.62; AUC=0.79) and VA loss (t=12.5, P<0.001; ES=0.64; SRM=1.48; AUC=0.78), and was somewhat responsive to lens opacity progression (t=4.58, P<0.001; ES=0.17, SRM=0.24, AUC=0.61). In a variance components model both progression to advanced AMD and VA loss contribute significantly to the variation in the mean difference in overall VFQ score. Progression of lens opacity did not make a statistically significant contribution. AMD and VA events contribute little to the variability of the general health or ocular pain subscales. Progression to advanced AMD contributes more than VA loss to the variation of subscale scores on all other subscales except general vision and color vision, where VA loss has a larger contribution. Conclusions: The NEI-VFQ is a responsive measure of vision targeted, health-related quality of life in persons who experience progression to advanced AMD or a VA loss ≥15 letters. The questionnaire was not as responsive to lens opacity events, which often occur without change in VA.
Keywords: age-related macular degeneration • cataract • visual acuity