Psychometric properties of PedsQL generic core scales for children with functional constipation in The Netherlands

Hartman, E. E., Pawaskar, M., Williams, V., McLeod, L., Dubois, D., Benninga, M. A., & Joseph, A. (2014). Psychometric properties of PedsQL generic core scales for children with functional constipation in The Netherlands. Journal of Pediatric Gastroenterology and Nutrition, 59(6), 739-747. DOI: 10.1097/MPG.0000000000000527

Objectives: The aim of the study was to evaluate the psychometric properties of the Dutch translation of the Pediatric Quality of Life Inventory version 4.0 Generic Core Scales (PedsQL) in children with functional constipation (FC). Methods: The PedsQL was completed by children with FC ages 5 to 18 years and by the parents of children ages 2 to 18 years. To assess construct validity, all of the parents and children completed the PedsQL Gastrointestinal Symptoms Module (GI module) and children ages 8 to 18 years completed the defecation disorder list. Item response distributions, internal consistency reliability, patient-parent agreement, and discriminating ability were evaluated. Results: Overall, 269 children were enrolled. The PedsQL showed minimal missing responses (self-report: 5.3% with >50% missing, parent report: 2.6% with >50% missing) and achieved satisfactory internal consistency for the total (self-report: alpha = 0.86, parent report: alpha = 0.88), physical health (self-report: alpha = 0.75, parent report alpha = 0.83), and psychosocial health (self-report: alpha = 0.80; parent report: alpha = 0.85) scores. Internal consistency was less convincing for those ages 5 to 7 years (alpha = 0.53-0.77) than for older individuals. Self-reported health-related quality of life was lower than parent reported, with the exception of children ages 5 to 7 years. Interrater reliability intraclass correlations were 0.52 to 0.61 for the overall population. Correlations among the PedsQL, GI module, and defecation disorder list were consistent with a priori hypotheses and generally supported construct validity. The PedsQL distinguished between a previously published healthy reference sample and children with FC, and between those with mild and severe symptoms (P < 0.05). Conclusions: The self-reported and proxy reported PedsQL demonstrated satisfactory measurement properties in Dutch children with FC ages 8 to 18 years. Further research is needed to establish internal consistency and validity in those ages 5 to 7 years