Timing and directions for administration of questionnaires affect outcomes measurement.

We used data from two pilot studies to compare the change in patients' self-reported health-related quality of life after participation in two nearly identical Department of Veterans Affairs (VA) Blind Rehabilitation Center (BRC) programs, the Southwestern BRC in Tucson, Arizona, and the BRC at the VA hospital in Hines, Illinois. Researchers at the Southwestern BRC administered the National Eye Institute Visual Functioning Questionnaire as directed by the developer. Researchers at the Hines BRC modified the directions to consider use of low-vision devices. Interval person-ability and item-difficulty measures estimated from patient responses pre- and postrehabilitation were compared with these same measures obtained at follow-up. At the Southwestern BRC, no change was reported in either person or item measures 3 months after rehabilitation. At the Hines BRC, improvement was seen in both the person and item measures when measurements were made immediately following rehabilitation. Because a temporary halo effect may explain the higher ratings at discharge, veterans from the Hines cohort were contacted by telephone and administered the same instrument 3 years later. For these subjects, the improvement noted in the person measure disappeared at follow-up, while the improvement in the item measure was maintained.