Refinement and validation of the ItchyQoL using classical test theory and item response theory resulted in a reduction of the response categories from a 5-point to a 3-point scale

Abstract

Background The ItchyQoL is an itch-specific patient-reported outcome measure used to assess quality of life in patients with chronic pruritus (CP). Objectives We aimed to assess and extend the psychometric properties of the Itchy- QoL using classical test theory (CTT) and item response theory (IRT). Methods Item characteristic curves were analysed to investigate whether the response categories were functioning optimally. Confirmatory factor analyses were carried out on the ItchyQoL prior to and after rescoring of the response categories. We conducted a Rasch analysis for the ItchyQoL with revised response options and assessed the mean fit residuals in addition to the assumptions of unidimensionality and local independence. Results In total, 551 patients with CP from nine European countries completed the 22 items of the ItchyQoL. IRT analysis supported the revision of response options from five points to three. This revision was supported by excellent structural validity using CTT. The overall fit to the Rasch model was adequate. Unidimensionality was supported by the ItchyQoL overall scale and by the single subscales; however, local independence was violated in eight cases. Conclusions We suggest a revision of the response categories of the ItchyQoL from a 5-point to a 3-point scale. When this revision was applied, the ItchyQoL showed excellent structural validity according to CTT and IRT/Rasch. The calculation of an overall ItchyQoL sum score is allowed.