What do different visual function questionnaires measure?

Purpose: There are dozens of visual function questionnaires (VFQs). VFQs differ in number of items, item content, and response categories. Despite differences, all VFQs are designed to measure vision-related functional ability in visually impaired people. The objective is to determine whether four popular VFQs measure the same variable in visually impaired patients. Methods: Two of four VFQs (ADVS, NEI-VFQ, VF-14, VAQ) plus the SF-36 were administered by telephone to 407 consecutive low vision clinic patients. Thus, each instrument was administered to just over 200 patients, and there were over 67 patients for each of the 6 pairings of VFQs. Separate Rasch analyses were performed on patient responses to each VFQ and to the physical and mental health domains in the SF-36. Results: Person measure estimates from the four VFQs were highly correlated. Person measures estimated from the mental health and physical domains of the SF-36 were uncorrelated with each other and with person measures estimated from the VFQs. From principal components analysis we concluded that three factors were necessary and sufficient: the first principal component accounts for 62% of the variance, the other two factors account for 16% and 14% of the variance, respectively. The VFQs load most heavily on the first factor; the mental health component of the SF-36 loads most heavily on the second; and the physical limitations component of the SF-36 loads most heavily on the third. Conclusions: The four VFQs measure the same vision-related functional ability variable in low vision patients that is separate from and independent of the physical and mental health variables measured by the SF-36.