Factors influencing self- and parent-reporting health-related quality of life in children with brain tumors

Abstract

Health-related quality of life (HRQOL) is not only a degree of health but also reflects patient perceptions and expectations of health. For children with brain tumors, better understanding of HRQOL requires the use of complementary reports from parents and interviewer-administered reports for children. Here, we aimed to test whether or not the trait anxiety of children and the psychological distress of their parents influence children's and parents' responses to HRQOL questionnaires, and whether or not the report-administration method for children influences children's responses to HRQOL questionnaires. One hundred and thirty-four children aged 5-18 with brain tumors and one of their parents completed the Pediatric Quality of Life Inventory(™) (PedsQL(™)) Brain Tumor Module questionnaires. In addition, the children also completed the State-Trait Anxiety Inventory for Children (STAIC), and the parents also completed the Kessler-10 (K10) and health and sociodemographic characteristics questionnaires. The child questionnaires were administered either by the child (self-administered) or an interviewer. Rater-dependent perceptions about HRQOL were derived from the subscales scores of the PedsQL(™) Brain Tumor Module using structural equation modeling based on a multitrait-multimethod model. The STAIC trait-anxiety score, K10 score, report-administration method, and other health and sociodemographic factors related to each child's or parent's perceptions were identified through multiple linear regression analyses of the questionnaire responses. We used a path analysis to estimate the change in a PedsQL(™) child-reported score that occurs when interviewer-administration changes the child's perception about HRQOL. Surveys for 89 children were self-administered while those for 45 were interviewer-administered. The perceptions of the children and parents were calculated by fitting data to the model (chi-squared P=0.087, normed fit index=0.932, comparative fit index=0.978, standardized root mean squared residual=0.053, and root mean square error of approximation=0.054). The children's perception of HRQOL was affected by their STAIC trait-anxiety score (b=-0.43, 95% CI [-0.60, -0.25]). The parent's perception was affected by their child's treatment status (b=0.26, 95% CI [0.09, 0.43]), the parent'sK10 score (b=-0.21, 95% CI [-0.37, -0.04]), and by education level (b=0.17, 95% CI [0.00, 0.34]). The change in the child-reported PedsQL(™) score in relation to the method of administration ranged from -1.1 (95% CI: -3.5, 1.3) on the procedural anxiety subscale to -2.5 (95% CI: -7.6, 2.6) on the movement and balance subscale. Child-reporting of HRQOL is little influenced by the method of administration. Children's perception about HRQOL tended to be influenced by their trait anxiety, while parents' perception was influenced by their psychological distress, academic background, and their child's treatment status.