No abstract

No abstract

No abstract

No abstract

Clinically useful and evidence-based mental health assessment requires the identification of strategies that maximize diagnostic accuracy, inform treatment planning, and make efficient use of clinician and patient time and resources. This study uses classification tree analyses to determine whether parent- and child-report instruments, alone or in combination, can accurately predict diagnoses as measured by the Anxiety Disorders Interview Schedule (ADIS). The ADIS, which is the gold-standard semistructured interview for anxiety disorders in children and adolescents, requires formal training and lengthy administration. Data were collected as part of the standard diagnostic assessment process for 201 patients (ages 5 to 17 years) in an urban outpatient psychiatry specialty clinic. Analyses examined 2 models to determine which predictors reached an acceptable level of diagnostic accuracy for generalized anxiety, social anxiety, and separation anxiety disorders. The first model used scores on a parent- and child-report anxiety measure combined with demographic factors, and the second model incorporated a broad-band measure of child psychopathology and a depression measure into the analysis. Although demographic factors did not emerge as accurate predictors in either model, particular measures, either alone or in combination, were able to predict specific ADIS diagnoses in some cases, allowing for the potential streamlining of ADIS administration. These results suggest that a classification-tree analysis lends itself to the construction of simple algorithms that have high clinical utility and may advance the feasibility and utility of evidence-based assessment strategies in real-world practice settings by balancing cost effectiveness, administration demands, and accuracy.