abstract
- PURPOSE: We estimated agreement in classifying psychiatric disorders using a structured interview and symptom checklist, examined associations between psychiatric disorder and health-related quality of life (HRQL) in children with chronic physical illness, and investigated if the instruments were statistically equivalent in predicting HRQL. METHODS: We analyzed data from children aged 4-16 years diagnosed with a chronic physical illness. Parents and children completed the Mini International Neuropsychiatric Interview for Children and Adolescents (MINI-KID; interview), and Emotional Behavioural Scales (EBS; checklist). Two approaches classified psychiatric disorder for the EBS: 1-thresholds based on prevalence (b-EBS); 2-T scores ≥ 70 (t-EBS). Health-related quality of life was measured using the KIDSCREEN-27. Instrument and informant agreement was estimated with kappa (κ). Regression models examined associations between psychiatric disorder at baseline and 24-month HRQL. The method of variance estimates recovery was used to determine whether the instruments were statistically equivalent. RESULTS: Agreement for any psychiatric disorder between the parent-reported MINI-KID and b-EBS was considered substantial (κ = 0.61), but moderate with the t-EBS (κ = 0.51). Fair agreement was found for child reports (κ = 0.28 and κ = 0.29, respectively). Parent-child agreement across instruments was low. No associations between psychiatric disorder measured by the MINI-KID vs. the b-EBS or t-EBS on child HRQL were significantly different for either informant. CONCLUSION: Agreement between instruments and informants was consistent with previous reports in other samples. Associations between psychiatric disorder and HRQL were not significantly different between the MINI-KID and EBS. Symptom checklists offer a practical opportunity to screen and monitor psychiatric disorder in children with physical illness.