Measuring health-related quality of life in children from different perspectives using the Pediatric Quality of Life Inventory™ (PedsQL™) and teachers' ratings

Background: Health-related quality of life (HRQL) is a concept being increasingly used in paediatric health care. Since health professionals have to rely on both child and adult reports of the child's condition, more information as to which factors moderate the correspondence between these reports would be useful. Aim and methods: This study attempted to identify the cross-informant variance (CIV) in child, parent and teacher reports about the child's HRQL in a large sample of elementary school children. We used the Austrian-German version of the generic core Pediatric Quality of Life Inventory™ (PedsQL™) for children and parents as well as a short HRQL rating scale derived from PedsQL™ for teachers. We identified the CIV with the use of intraclass correlation and paired score differences in 620 triple assessments, compared them to results of other PedsQL™ studies and tried to identify the factors having an impact on CIV and on the self-reported HRQL of children whose proxy ratings demonstrated a large CIV. Results: On the whole, our results indicated low to medium agreement between children and parents, similar to results of other PedsQL™ studies in schoolchildren. The only variable that was found explaining child-parent variance was the child's gender, with girls deviating more from their parents than boys. Child-teacher and parent-teacher CIV seemed to be primarily an effect of the child's school performance and the teacher-parent relationship. Children whose proxy ratings differed by more than ± 15 scores showed significantly lower self-reported HRQL in some areas as compared to children whose proxy ratings showed less CIV. Conclusion: Our results underline that it is more difficult to get concordant HRQL ratings of schoolchildren as compared to children with chronic disease. © 2008 Springer-Verlag.