Purchase this article with an account.
Judith Goldstein, Eva Fenwick, Robert Patrick Finger, Vijaya Gothwal, Mary Lou Jackson, Ecosse Luc Lamoureux, Gwyneth Rees, Robert W Massof; Calibration of the Impact of Vision Impairment Instrument (IVI) Questionnaire. Invest. Ophthalmol. Vis. Sci. 2018;59(9):3913.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Provide calibrated item estimates for the Impact of Vision Impairment instrument (IVI) to improve uniformity in measurement across studies. Working collaboratively and pooling data, our aim is to maximize precision in these item estimates, such that they can function as an item bank. This item bank will be available to researchers and clinicians to support measurement of visual ability in people with vision impairment.
Eight datasets from 6 principal investigators representing responses to IVI items from 2867 low vision patients were pooled for analysis. Eligible patients were ≥18 years and from Australia, India, and the United States. Data from each sample included raw scores of patient responses to the IVI, visual acuity (VA), ocular diagnosis, date of birth or age, and sex. Rasch analysis, using the Andrich Rating Scale Model (Winsteps ver 3.65), was performed on pre-intervention IVI responses to estimate person and item reliability, item measures and response category thresholds, and a set of person measures for each patient. “Emotional Well-being” (8 items) were excluded in the estimates of the person measures as Lamoureux, et al., have shown that these items represent a separate measure. Differential item functioning (DIF) by patient sample was evaluated.
Patients ranged in age from 18 to 103 years with a median age of 62, with males representing 55% of the sample. Macular degeneration was the primary diagnosis in 29% of all patients, while the residual represented all other ocular disorders. Visual ability as estimated by person measures from the IVI scaled monotonically with VA. Mean person measures for patients ≥20/40 (n=414) were 0.51 (SD 1.11), 0.16 (SD 1.04) for patients <20/40 to 20/60 (n=467), 0.09 (SD 0.96) for patients <20/60 to 20/200 (n=984) and -0.23 (SD 0.92) for patients <20/200 (n=995); VA was missing in 7 patients. There were indications of non-uniform DIF in one dataset. In examining covariates, this dataset contained subjects that were younger and had worse visual acuity than patients in the other datasets.
Item difficulty estimates from a varied dataset of people with visual impairment supports the ability to calibrate IVI measures and shows minimal non-uniform DIF. This item bank may be used by researchers and clinicians for comparative estimates of functional ability in a common unit of measurement when studying baseline traits or treatment outcomes.
This is an abstract that was submitted for the 2018 ARVO Annual Meeting, held in Honolulu, Hawaii, April 29 - May 3, 2018.
This PDF is available to Subscribers Only