Abstracts

Rationale: Children with chronic conditions are increasingly being invited to provide information about outcomes that can only be obtained through self-report such as quality of life (QOL) or its predictors (e.g., mental health (MH)). These reports are often complementary to, but occasionally different from, those of parents. To understand child and parent report, we sought to determine: i) whether one might expect systematic differences for certain or all QOL and MH constructs; ii) if reporting differences were indicative of stable child-parent dyad subgroups; and iii) what traits characterized membership in one type of dyad group (i.e., ‘child positive-parent negative’, ‘parent positive-child negative’ or ‘child-parent agreement’).Methods: We obtained two reports, a mean of 8.2 months apart, from 480 Canadian dyads involving children with epilepsy aged 8-14 and a parent, using an epilepsy-specific measure (CHEQOL), a generic measure of QOL (KIDSCREEN 52); and validated standardized measures of child MH components of attention, conduct, cooperativeness and depression (from Canadian Population Health Scales). Each QOL subscale or MH component total parent score was deducted from the respective child score and displayed as a box plot at each time point. Membership in discrepancy groups was based on score differences greater than 1 SD away from the mean. Stability of membership in a dyad group was determined across time and across constructs with iterative frequency counts resulting in stability scores from 0-1. Predictors of membership in one type of dyad group for MH and QOL constructs was were determined using a logistic regression approach.Results: Overall, children were significantly more likely to report positive overall life satisfaction, positive friendships, lower depression scores, and higher parental supportiveness than their parents (p<0.05). Remaining QOL and MH domains reflected overall agreement for most dyads in the sample. For example, for epilepsy specific QOL 49 % of the dyads demonstrated a perfect stability score of 1and another 22% having a stability score of 0.75. Maternal education and seizure severity were the most common significant predictors of membership in the ‘child positive-parent negative’ group.Conclusions: Certain constructs within QOL and MH reporting (e.g., depression) can be expected to be discrepant between parents and children in the 8-14 year age range. This study demonstrates that these reporting discrepancies are authentic and stable characteristics of the parent-child dyad even when children and parents disagree. Additional information predicting membership in a dyad group can provide clinicians with information about possible explanations for reporting discrepencies.