Factors influencing agreement between child self-report and parent proxy-reports on the Pediatric Quality of Life Inventory™ 4.0 (PedsQL™) generic core scales

BACKGROUND In situations where children are unable or unwilling to respond for themselves, measurement of quality of life (QOL) is often obtained by parent proxy-report. However the relationship between child self and parent proxy-reports has been shown to be poor in some circumstances. Additionally the most appropriate statistical method for comparing ratings between child and parent proxy-reports has not been clearly established. The objectives of this study were to assess the: 1) agreement between child and parent proxy-reports on an established child QOL measure (the PedsQL) using two different statistical methods; 2) effect of chronological age and domain type on agreement between children's and parents' reports on the PedsQL; 3) relationship between parents' own well-being and their ratings of their child's QOL. METHODS One hundred and forty-nine healthy children (5.5 - 6.5, 6.5 - 7.5, and 7.5 - 8.5 years) completed the PedsQL. One hundred and three of their parents completed these measures in relation to their child, and a measure of their own QOL (SF-36). RESULTS Consistency between child and parent proxy-reports on the PedsQL was low, with Intra-Class correlation coefficients ranging from 0.02 to 0.23. Correlations were higher for the oldest age group for Total Score and Psychosocial Health domains, and for the Physical Health domain in the youngest age group. Statistically significant median differences were found between child and parent-reports on all subscales of the PedsQL. The largest median differences were found for the two older age groups. Statistically significant correlations were found between parents' own QOL and their proxy-reports of child QOL across the total sample and within the middle age group. CONCLUSION Intra-Class correlation coefficients and median difference testing can provide different information on the relationship between parent proxy-reports and child self-reports. Our findings suggest that differences in the levels of parent-child agreement previously reported may be an artefact of the statistical method used. In addition, levels of agreement can be affected by child age, domains investigated, and parents' own QOL. Further studies are needed to establish the optimal predictors of levels of parent-child agreement.