CHICAGO—A review of diagnostic instruments for pediatric bipolar disorder (PBD) revealed considerable variation in administration, scoring, and DSM-IV adherence, Prudence Fisher, PhD, stated in a presentation at the 55th Annual Meeting of the American Academy of Child and Adolescent Psychiatry.

Dr. Fisher, Assistant Professor of Clinical Psychiatric Social Work, and coauthor Cathryn Galanter, Assistant Professor of Clinical Psychiatry, both in the Division of Child and Adolescent Psychiatry at Columbia University in New York City, and their colleagues reviewed the following diagnostic interview instruments: Kiddie Schedule for Affective Disorders and Schizophrenia-Epidemiological version (KSADS-E), KSADS-Present and Lifetime version (KSADS-PL), Washington University KSADS (WASHU-KSADS), Child and Adolescent Psychiatric Assessment (CAPA), Diagnostic Interview Schedule for Children Version IV (DISC-IV), and the Missouri Assessment of Genetics Interview for Children (MAGIC). These tools were identified as the most commonly used, following a PubMed search. Reviews of informants, interviewer training, scoring and combining methods, and adherence to DSM-IV criteria were scrutinized, "for example, what symptoms were elicited to determine presence or absence of a criterion, whether presence of a criterion required behavior changes from baseline in all symptoms, and whether the instrument required that the B-criteria co-occur with the mood episode (A-criteria) for all criteria," Dr. Fisher commented.

KSADS-E obtained data from children if they were between the ages of 12 and 17, and DISC-IV from children 9 and older, but both obtained parent data in all cases. All other instruments obtained data from both child and parent during separate interviews. The interviewers differed as well; KSADS-E, KSADS-PL, and WASHU-KSADS were designed to be administered by clinicians, but there were several instances of trained laypeople obtaining the data. The other tests were designed to be given by laypeople.

Use of the Likert scale differed among instruments. KSADS-E, KSADS-PL, CAPA, and MAGIC used the 3-point scale, while WASHU-KSADS used the 6-point scale. Which number on the scale qualifies as meeting a symptom threshold varied among all instruments. DISC-IV used the dichotomous (yes/no) scale. The combination of informants also differed among groups. WASHU-KSADS and DISC-IV labeled data from either informant as sufficient, while CAPA and MAGIC developed separate symptom scores for each (KSADS-E analyses were unavailable).

The biggest differences between groups were seen in the examination of diagnostic algorithms. Diagnoses were based on a manual scoring sheet and a consensus conference of clinicians in both the KSADS-E and WASHU-KSADS assessments. MAGIC used the manual scoring sheet only. CAPA and DISC-IV produced diagnoses on the basis of a computerized scoring algorithm. KSADS-PL used clinical judgment without a manual scoring sheet.

The questions and specific symptoms used to assess PBD criteria differed. Some WASHU-KSADS groups considered only elated mood as PBD criterion. WASHU-KSADS and CAPA used irritable mood gathered from depression module questions. Many differences were seen in the assessment of goal-oriented behavior and pleasurable and risky activity. The way criteria were measured from baseline also differed. KSADS-PL, DISC-IV, and MAGIC assessed all behavior changes from baseline. Four out of seven symptoms were accepted by WASHU-KSADS and CAPA. All instruments required that B-criteria co-occur with the mood episode, except KSADS-PL, which considered five out of the seven symptoms.

"Child psychiatry would benefit from a more uniform method of assessing and evaluating phenomenology and determining diagnoses in PBD," Dr. Fisher said. Variation in administration, scoring, and DSM-IV adherence could account for differences in phenomenology, family history, comorbidity, pathophysiology, and treatment response.

—Laura Sassano

Return to table of contents