Comparing the measurement properties of visual analogue and verbal rating scales.

Abstract

Utilising Rasch analysis on the Keratoconus Outcome Research Questionnaire (KORQ) data, we explored the hypothesis that the KORQ with discrete verbal rating scale (VRS) would demonstrate better psychometric properties and provide less noise in measurement than with a visual analogue scale (VAS). The KORQ is a keratoconus-specific patient-reported outcome measure; it has activity limitation and symptoms scales. The KORQ scales with two different rating scales (VAS and a discrete 4-response VRS) were completed by self-administration by people with keratoconus. For each KORQ scale, Rasch analysis-based psychometric properties were compared between the two versions. Rasch analysis was also used to optimise rating scale functioning when disordered thresholds were observed. 118 (mean age±SD, 46.4±0.4years) and 169 (45.4±14.7years) people completed the KORQ with VAS and VRS, respectively. Both scales demonstrated high measurement precision. However, the VAS rating scale was disordered (6 out of 11 categories dysfunctional) and had two misfitting items. Conversely, the VRS had ordered categories and no misfitting items. For the disordered VAS, ordering was achieved only after collapsing 11 categories into four categories. In comparison to the KORQ with VRS, the repaired VAS had lower measurement precision, test information, variance explained by the measure, poor targeting, and reduced measurement range. The KORQ demonstrated superior psychometric properties when measured using a VRS than with a VAS. This illustrates the advantages of verbal rating scales for a patient-reported outcome measurement over a visual analogue scale.