Comparing diagnostic checklists for pediatric bipolar disorder in academic and community mental health settings

Objectives: To compare six promising mania measures, the Parent Mood Disorder Questionnaire (P-MDQ), the Adolescent self-report MDQ, the 10-item short form of the Parent General Behavior Inventory (PGBI-SF10), the 28-item Adolescent General Behavior Inventory (AGBI), the Parent Young Mania Rating Scale (P-YMRS), and the adolescent YMRS, in a demographically diverse outpatient sample. Methods: Participants were 262 outpatients ( including 164 males and 131 African-Americans) presenting to either an academic medical center (n = 153) or a community mental health center (n = 109). Diagnoses were based on semi-structured interviews with the parent and then youth sequentially. Results: Ninety youths (34%) met criteria for a bipolar spectrum disorder. Parent measures yielded Areas Under the Receiver Operating Curve (AUROC) values of 0.81 for the PGBI-SF10 to 0.66 for the P-YMRS. Adolescent report measures performed significantly less well, with AUROCs ranging from 0.65 to 0.50. There were no significant differences in the diagnostic performance of the measures across the sites or by racial groups, although the reliability of measures tended to be lower in the urban community mental health site. The PGBI-SF10 made a significant contribution to logistic regression models examining all combinations of the instruments. The P-MDQ added information in the younger age group, and no measure improved classification of bipolar cases after controlling for the PGBI-SF10 in the older age group. Discussion: Results replicate previous findings that, in decreasing order of efficiency, the PGBI-SF10, P-MDQ, and P-YMRS significantly discriminate bipolar from non-bipolar cases in youths aged 5-18; and they appear robust in a demographically diverse community setting. Adolescent self-report measures are significantly less efficient, sometimes performing no better than chance at detecting bipolar cases.