Validity and reliability of the Turkish version of the KIDSCREEN-27 for individuals with cerebral palsy

Abstract

Our aims in this study were to examine the construct/concurrent validity and internal/test-re-test reliabilities of both the self-report and parent-report questionnaires of a Turkish version of the KIDSCREEN-27 for adolescents with cerebral palsy (CP). We used a convenience sample of 135 children and adolescents with CP aged 8–18 years (65 males, 70 females; M age = 12.39, SD = 3.57) and their parents/caregivers (123 mothers, seven fathers, and five grandmothers). We explored structural construct validity via confirmatory factor analysis (CFA). Concurrent validity was examined via Spearman’s correlations between the KIDSCREEN-27 questionnaires and the self-report and primary caregiver report forms of the Cerebral Palsy Quality of Life for Children (CP QOL-Child) and adolescents (CP QOL-Teen). We explored test-retest and internal consistency reliabilities utilizing intraclass correlation coefficients (ICC) and Cronbach’s alpha (α), respectively. CFA goodness-of fit indices verified that the predefined model of the KIDSCREEN-27 was a good fit for data from the CP population (X2/df<5, GFI >0.90, AGFI >0.90, RMSEA<0.80). Results showed the subdomains of both KIDSCREEN-27 forms to be significantly correlated with the matched subdomains of the CP QOL-Child and CP QOL-Teen questionnaires (ranges of r = 0.62–0.94). Overall, α and ICC coefficients for all subdomains of both KIDSCREEN-27 versions were acceptable (α and ICC >0.70). Both versions of the KIDSCREEN-27 have satisfactory psychometric properties for use in evaluating health-related quality of life (HRQOL) outcomes in children and adolescents with CP.