National Eye Institute Visual Function Questionnaire-25 (NEI VFQ-25) Calibrations for Low Vision Patients with age-related macular degeneration (AMD)

It is recognized in ophthalmology that changes in visual acuity do not entirely illustrate the effects of an intervention, and with increasing options for pharmaceutical therapy in retinal disorders, an understanding of the benefits from the patient perspective must be clearly understood.

Evaluation of visual ability, the ability to perform activities that depend on vision, can be made quantitative for measuring outcomes from the patient’s perspective by employing a visual function questionnaire (VFQ). Most VFQs have a fixed set of items that describe specific vision-dependent
daily activities and ask the patient to rate the difficulty of performing those activities using an ordered set of response categories. The 25-item National Eye Institute Visual Functioning Questionnaire (NEI VFQ-25), is one instrument used, however, it remains difficult to compare these findings across
studies, as mean changes in raw scores are often published. The non-linearity of raw scores creates challenges to the meaningfulness of the findings, and raw scores are instrument-specific, which makes generalizing results difficult. The NEI VFQ-25, like most VFQ’s, permits opt out responses to
irrelevant items (e.g., “not important”, etc.) which are scored as missing data. This reduces measurement precision, and depending on the estimation method employed, can distort the visual function measure. Rasch analysis can be employed to address both the issues of missing data and nonlinearity
of raw scores. This is best achieved by calibrating tem difficulty on the NEI VFQ-25. Item calibration then positions us to employ computer adaptive testing when administering the NEI VFQ-25.

Our aim to calibrate items on the NEI VFQ-25 involves assigning a difficulty estimate to those items that have a rating scale. Our group has performed this type of analysis with the Activity Inventory VFQ using 3200 low vision patients to calibrate items. To maximize precision of item calibrations, we
will want to maximize the number of patient responses to the NEI VFQ-25. Our team has collected responses to the NEI VFQ-25 from approximately 1000 patients from low vision clinics across the U.S. With additional data from Genentech studies we can improve our precision of the estimates while
at the same time ensuring that the item calibrations are relevant to patients participating in retinal related
clinical trials. We anticipate that this information may be of benefit to Genentech/Roche as it pursues various presentations and publications of its scientific data involving NEI VFQ-25 information, and may be of value in its indication pursuits with the EMA, FDA or other regulatory bodies that might be considering vision-related PRO’s. Once analysis is complete, our plan is to publish item calibrations in the ophthalmic literature to enhance future comparative research using publically available NEI VFQ-25 item calibrations.