Cataract Symptom Score Questionnaire: Rasch Revalidation

Abstract

Purpose: The 5-item Cataract Symptom Score (CSS) was developed using classical test theory to determine appropriate timing for cataract surgery, its outcomes, and whether the symptoms included bother cataract patients. The purpose of this study was to investigate the psychometric properties of the 5-item CSS using Rasch analysis. Methods: Responses from 113 patients awaiting cataract surgery to the 5-item CSS (bothered by double or distorted vision, glare or halos, blurry vision, colors looking different and worsening of vision) were subjected to Rasch analysis. The use of response categories (threshold order), ability of CSS to discriminate between participants' severity of symptoms (person separation, recommended minimum value 2.0), matching of item difficulty to severity of symptoms (targeting), and ability of items to measure a single construct (unidimensionality) were investigated. Results: Participants used the response categories as intended. However person separation (1.74) was inadequate, suggesting that the CSS could differentiate only between two groups of participants by their symptoms. Furthermore the CSS was poorly targeted for our population, as the majority of our participants were not bothered by symptoms in the CSS. All items fit the single construct, implying that the CSS is a unidimensional measure of cataract symptoms. Conclusions: The CSS is unable to discriminate people with cataract. This problem could be fixed by adding additional questions, but a superior approach may be to create an item bank of cataract symptoms questions, including those of the CSS, and utilize computer-adaptive testing for measurement.