Validity, reliability and dimensionality of the Norwegian version of Wound-QoL

Abstract

The aim of the study was to explore the psychometric properties of the Norwegian version of the Wound-quality of life (QoL)-17. We included 204 patients with hard-to-heal wounds on the lower extremity. Patients filled out the Wound-QoL-17, SF-36 and Patients' Global Impression of Change at three time points over 14 weeks. Clinical and demographical data were collected at baseline. Wound severity was collected at baseline and 14 weeks follow-up (T2). Confirmatory factor analyses showed acceptable fit of the hypothesised three-factor model (i.e., body, psyche and everyday life) of the Wound-QoL as long as we accepted two correlated error terms within the Body factor (χ² = 203.14, p = 0.000, df = 99, χ²/df = 2.05, Root Mean Square Error of Approximation = 0.072, Standardised Root Mean Square Residual = 0.059, Comparative Fit Index = 0.943 and Tucker–Lewis Index = 0.930). Correlation showed moderate to strong associations between Wound-QoL and SF-36 (−0.400** to −0.777**), and significant associations between Wound-QoL and Patients' Global Impression of Change (0.199*), general wound pain intensity (0.435**), pain intensity at wound change (0.340**) and infection (0.174*). The intraclass correlations, ranging from 0.578** to 0.782**, suggested strong test–retest reliability. Cronbach's alphas for all subscales and the total score between 0.748 and 0.922, indicated good internal consistency. The Norwegian version of Wound-QoL demonstrates good reliability and construct validity and is suitable for evaluating QoL in patients with hard-to-heal wounds. However, some improvements were performed to achieve an acceptable fit.