The NEI VFQ 51-item test version and its relationship with visual acuity across 5 diseases

Academic Article


  • Purpose: To determine the psychometric properties of the NEI-VFQ, a survey designed to assess visual functioning and the influence of visual disability on health-related quality of life (HR-QOL). Methods: We conducted a 5-site, cross-sectional study that included an interview with the 51-Item NEI-VFQ, medical history, and demographic variables. Eligible persons with one or more of the following conditions: cataract (n=81); glaucoma (n=72); diabetic retinopathy (n=58); age-related macular degeneration (n=56); or CMV retinitis (n=21) were identified by chart review. Internal consistency of the NEI-VFQ was estimated with Cronbach's coefficient α. Correlations between logmar-transformed visual acuity in the better and worse eye and the 10 NEI-VFQ sub-scale scores were calculated. Results: Overall, 262 patients (mean age 61±18 years, 46% female) were interviewed. Median visual acuity (VA) in the better eye was 20/30 (range: 20/15 to 20/400) and in the worse eye was 20/50 (range: 20/20 to 20/400). Mean±SD NEI-VFQ scores (0-100 scale, best=100) included: near vision: NV= 71±27, distance vision: DV=71±25, general vision: GV = 63±21, limitations in social functioning due to vision: SF=83±22, role limitations due to vision: RL=81±24, dependency due to vision: DEP=81 ± 24, mental health symptoms due to vision: MH=68±23, future expectations for vision: EXP=51 ± 26, driving difficulties: DR=76±21, and ocular pain: OP=88±16. Cronbach's coefficient a for each sub-scale ranged between 0.94 and 0.74. Correlations (Pearson's R) between VA in the better eye and NEI-VFQ sub-scales were (p<.05 for all correlations): NV DV GV SF RL DEP MH EXP DR OP VA: -0.53 -0.51 -0.44 -0.43 -0.38 -0.35 -0.27 -0.20 -0.19 -0.14 As was expected, correlations between VA in the worse eye and the NEI-VFQ scores were lower but remained significant. Persons who reported less involvement in activities that require vision over the previous week were more likely to report poorer NEI-VFQ scores. Conclusion: The NEI-VFQ detected disability across all conditions. Although on average, the participants had good VA, they reported moderate levels of limitations across all dimensions of visual functioning and HR-QOL assessed by the NEI-VFQ. The dimensions of functioning that specifically assess performance of visual activities (NV and DV) were most highly correlated with VA in the better eye. Sub-scales that assess the impact of visual disability on well-being and mental health (MH and EXP) had lower correlations with VA. The high internal consistency suggests that the NEI-VFQ Test Version can be shortened while retaining adequate reliability.
  • Authors

    Author List

  • Mangione CM; Lee PP; Berry SH; Spritzer K; Janz NK; Klein R; Owsley C; Hays RD
  • Volume

  • 37
  • Issue

  • 3