Abstracts

Rationale: Youth with epilepsy (YWE) have increased risk for psychosocial comorbidities. We aim to support identification and management of comorbidities through a screening program in a pediatric epilepsy clinic. There are several screening measures available that assess psychosocial challenges in youth. Here, we assess which screening tools are most appropriate by 1) comparing the sensitivity and agreement of two commonly used parent-report emotional screening measures and 2) examining agreement between parent and child report.

Methods: Participants were 20 YWE ages 6.3 – 17.0 (M = 12.0 years) and their parents. All parents completed screening measures and YWE ages 11 – 17 completed screening measures (n = 10). Although part of a larger study, we focus on measures of mood and anxiety. Parents completed the Pediatric Quality of Life Inventory-Epilepsy Module (PedsQL) and the Strengths and Difficulties Questionnaire Parent Report (SDQ-P). Youth completed the Strengths and Difficulties Questionnaire Child Report (SDQ-C). Emotional subscales consisted of 5 items for both measures (PedsQL Mood/Behavior subscale and SDQ Emotional Symptoms subscale). Raw scores and whether a rating met clinical cutoff for a positive screen (i.e. at risk) were analyzed descriptively and using Cohen’s kappa coefficient and point-biseral correlation.  

Results: The PedsQL and SDQ-P demonstrated minimal agreement (κ = .29) for cutoff ratings. Parent ratings met the clinical cutoff for 12 (60%) YWE on the PedsQL of which, 4 YWE also met criteria on the SDQ-P. The SDQ-P did not have any YWE that met clinical cutoff not already identified by the PedsQL. Two (20%) child ratings met clinical cutoff on the SDQ-C, and their parents’ ratings also met clinical cutoff. Of the remaining 8 child ratings that did not meet the cutoff, 4 (40%) had a parent that also did not meet cutoff on their ratings, and 4 (40%) had a parent rating that met clinical cutoff. Given that child ratings had few positive screens, raw scores from the child SDQ were examined; however, SDQ-C raw scores were not correlated with whether a YWE had a parent rating that met the clinical cutoff or not (r = 0.18, p = .61). 

Conclusions: A majority of YWE were at risk for mood and anxiety. We found poor agreement between the PedsQL emotional subscale vs. the SDQ-P emotional subscale. The PedsQL identified more YWE at risk and the SDQ-P did not identify any unique cases of at risk YWE. Our results suggest that the PedsQL is the better measure when the cost of missing a YWE that is at risk is higher than referring someone that may not need treatment. Our YWE self-report results support prior findings that youth commonly underreport emotional challenges. Conducting screening in a fast-paced epilepsy clinic is important and our results suggest that the PedsQL emotional subscale is a promising tool for screening emotional comorbidities in YWE given its sensitivity and brevity. Future work will examine additional screening measures that assess other areas of psychosocial challenges.

Funding: Hess Foundation