PURPOSE: To test the validity of the Impact of Visual Impairment (IVI) questionnaire in a cataract population.

METHOD: Cataract patients recruited from a hospital waiting list completed the IVI questionnaire. The scale was assessed for fit to the Rasch model. Unidimensionality, item and person fit to the model, response category performance, differential item functioning (whether different subgroups responded differently), and targeting of item difficulty to patient ability were assessed.

RESULTS: Overall, the IVI questionnaire performed well; there were ordered thresholds, person separation reliability was 0.97, and it was free from differential item functioning. One item (worry about eyesight getting worse) misfit the model and was removed. There was evidence of multidimensionality, indicating that the overall IVI score should be discarded; however, the 3 subscales (reading and accessing information, mobility and independence, and emotional well-being) functioned well. Several items calibrated differently in cataract patients compared with low-vision patients, indicating different issues are important to each population and that there is a need for population-specific conversion algorithms. Targeting of the IVI items was biased toward more impaired patients.

Figure 2. The targeting map shows an uneven spread of items across the full range of respondents’ scores suggesting less than optimal targeting of the cataract patients (top) to the IVI items and thresholds (bottom)..

CONCLUSIONS: The 3 subscales of the IVI questionnaire functioned well in a cataract population. However, additional items targeting the less impaired patients, especially second-eye cataract patients, would improve measurement.

    IVI cataract surgery   as PDF (672 Kb)

Index of Papers