Reliability and validity of the child health questionnaire-child form (CHQ-CF87) in a Dutch adolescent population.

Abstract

Feasibility, reliability, and discriminative validity of the cross-culturally adapted Dutch version of the originally US child health questionnaire-child form (CHQ-CF87), an 87-item generic pediatric health-related quality of life instrument, were assessed. The success criterion in this first evaluation was the equivalence of psychometric properties of the adapted and the original CHQ. A total of 466 schoolchildren (9-17 years) were invited to complete the questionnaire in the classroom. Test-retest reliability was measured after 14 days in a subgroup (n = 71). Response was 96%. Four scales had ceiling effects (>50%), as was reported in an Australian study. Cronbach alpha-values were adequate (>0.70), except for 'physical functioning' (0.56). Test-retest correlations, not previously reported, were not statistically significant for two CHQ-scales, whereas average retest scores indicated better health for five scales (p < 0.01). The CHQ scales discriminated significantly (p < 0.01) between children without (n = 281) and children with two or more self-reported chronic diseases (n = 59). This is in correspondence with US and Australian reports. The current data support application of the Dutch CHQ-CF in predominantly healthy populations, e.g. in school settings. Given the limitations of this study and some less favorable results (score distributions, internal consistency, test-retest reliability), further evaluation of the CHQ-CF is recommended, preferably by analyses of item performance and scale validity in international data sets that include varied clinical subgroups.