Abstract
Purpose: :
Many questionnaires have been developed to measure visual disability (visual functioning). However, simple summary scoring of ordinal responses limits their measurement properties including interchangeability of scores across instruments. The aim of this project was to use Rasch analysis to calibrate all items from all instruments onto a single linear scale of visual disability.
Methods: :
People on the waiting list for cataract surgery at Flinders Medical Centre, South Australia, were invited to participate in the study. Each participant completed, by self-administration, a number of the 19 instruments used in this study. In all cases, the Visual Disability Assessment (VDA) instrument was completed and used to anchor item difficulty across multiple analyses. The other instruments included the ADVS, VF-14, NEI-VFQ, VAQ, Catquest, IVI, TyPE, VCM1, VSQ, CSS, VFQOL, QOLVFQ, AVL, HVAT, VFI, DVI and ICS. A total of 353 items were calibrated using Rasch analysis (Winsteps v3.67); the performance of the Rasch model was fully evaluated. Comprehensive demographic data were collected.
Results: :
566 participants were recruited: 44% were male, 56% female; mean ± SD age was 74.1 ± 9.4 years; 59% had bilateral cataract, 41% were awaiting second eye surgery; 48% had co-morbid eye disease (age-related macular degeneration and glaucoma 16% and 9%, of total respectively); on average participants had been diagnosed with cataract for 3.2 ± 8.7 years and visual acuity was R 0.22 ± 0.20 (6/9.5-) in the better eye and L 0.55 ± 0.36 (6/24++) in the worse eye. Rasch estimates of item difficulty were generated with an average standard error on the measure of 0.11 logits indicating a high level of stability across items. 21.2% of items mis-fit the model. These misfitting items appear to tap a quality of life trait other than visual disability with evidence of multidimensionality on principal components analysis of the residuals. On the whole, item difficulty was well targeted to participant ability (difference between person and item means 1.01 logits), with a preponderance of items to measure the less able.
Conclusions: :
353 items have been calibrated on a single scale of visual disability thus forming an item bank. Items can be drawn from the bank for use in their native instruments, new instruments, or computer adaptive testing. All three approaches measure visual disability on a common scale, facilitating worldwide comparison of results across instruments and studies.
Keywords: quality of life • clinical (human) or epidemiologic studies: biostatistics/epidemiology methodology • treatment outcomes of cataract surgery