OBJECTIVES: The Dermatology Life Quality Index (DLQI; 10 items) is a generic dermatology health-related quality of Life (HRQoL) measure that is the most commonly used in dermatology. Despite its popularity little research has been conducted into the dimensionality of the questionnaire. The purpose of this study was to examine its scaling properties and establish whether it is unidimensional.
METHODS: DLQI data were combined from two studies; one involving people with psoriasis and the other patients with atopic dermatitis. Item Response Theory was used to determine; overall fit to the Rasch model, individual item fit, targeting of scale to severity of respondents, functioning of response categories and the presence of Differential Item Functioning (DIF) by disease, age or gender.
RESULTS: The sample included 146 psoriasis patients (male 50%, mean age 44.2 range 17–83 years) and 146 atopic dermatitis patients (male 50%, mean age 45.5, range 20–82 years). The DLQI misfit the Rasch model (Chi2 63.38, df 40, p 0.01). Item 2 misfit the Rasch model and items 5 and 7 showed borderline misfit. Items 4, 6, 7, 8 and 9 had disordered response thresholds indicating that these did not work in a logical way. Results showed a lack of spread in the measurement of HRQoL with too few items covering either milder or more severe levels of HRQoL. DIF by disease was shown in items 4 and 7 and DIF by age in item 10. After removal of item 2 and rescoring the response categories the DLQI still misfit the Rasch model (Chi2 54.92., df 36, p 0.02).
CONCLUSIONS: The results of the Rasch analysis showed there were several problems with the scaling properties of the DLQI and that little confidence can be placed in raw scores generated from the scale. These problems need to be addressed before the QLDS can be considered a valid and useful outcome measure.
