Reliability and Validity of the Children's Interview for Psychiatric Syndromes-Parent Version (P-ChIPS) in Youngsters with Autism Spectrum Disorders

Background: Research has shown that youngsters with ASDs present with high rates of behavior and emotional problems, including tantrums, mood swings, aggression, self-injury, and irritability. Psychiatric disorders are also quite prevalent; the most commonly-reported are disruptive behavior, mood, and anxiety disorders. Researchers have begun to examine psychiatric disorder in ASD using structured interviews. However, little research has examined the reliability and validity of these instruments, which are considered the gold standard in psychiatric research. One such interview is the Children’s Interview for Psychiatric Syndromes-Parent Version (P-ChIPS). It is a structured interview designed to concisely assess psychopathology according to DSM-IV criteria in clinical and epidemiological research with children and adolescents 6 to 18 years old. 
Objectives: This study examined the reliability and validity of the P-ChIPS in children and adolescents with ASDs. We investigated the agreement between the P-ChIPS and the Child and Adolescent Symptom Inventory (CASI; a parent-completed DSM-IV-based measure of psychiatric disorders) on the assessment of mood, anxiety and disruptive behavior disorders. We will also measure internal consistency and assess interrater reliability.

Methods: The parents of 60 children and adolescents (mean=11.2±3.8 years; range 6-17) with autism, Asperger’s disorder, and PDD-NOS were interviewed with the P-ChIPS and completed the CASI rating scale. The youngsters were administered the Stanford-Binet Intelligence Scales-Fifth Edition, and all had IQs>40. Diagnosis of ASD was confirmed by the Autism Diagnostic Interview-Revised. Agreement between the P-ChIPS and CASI was assessed with kappa statistics and overall agreement. The impact of IQ (IQ>70 vs. IQ<70), age (6-11 vs. 12-17 years), and language (verbal vs. nonverbal) on agreement was also examined by calculating kappa coefficients separately and comparing with z tests.
Results: Diagnostic agreement between P-ChIPS and CSI was higher for the disruptive behavior disorders and for major depression/dysthymia. Kappa statistics range from .20-.56 for the disruptive disorders, indicating fair to moderate agreement. Overall agreement ranged from 65-90%.  Kappa statistics for the mood disorders were .35 and .22, indicating fair agreement. Overall agreement was 83% and 78%.  Agreement was lower for the anxiety disorders, with most kappa coefficients indicating poor agreement (range .12-.48) and overall agreement ranging from 62%-80%. Age and IQ did not significantly influence kappa statistics.

Conclusions: Kappa statistics and overall agreement was lower than values reported in non-ASD samples. Age and IQ did not appear to impact kappa values. The P-ChIPS shows promise in the population, but may require adaptations for the ASD population.