Volume 22, Issue 1 (1993)
Conditional Probabilities in the Diagnosis of Depressive and Anxiety Disorders in Children
Jeff Laurent, Steven Landau, Kevin D. Stark
Differential diagnosis of anxiety and depressive disorders in children has been an area of confusion and controversy. A prototypic model of classification, which allows for heterogeneity of diagnostic criteria, would seem more appropriate when studying these disorders. Consistent with this model, a heuristic procedure using conditional probabilities is presented to determine the most efficient inclusion and exclusion criteria for the differential diagnosis of depression and anxiety in children. A multistage selection procedure was used to identify subjects from an initial pool of 720 fourth through seventh graders. Ultimately, 34 students were diagnosed as depressed, 30 as anxious, and 70 students did not receive either diagnosis. Base rates, sensitivity, specificity, positive predictive power, and negative predictive power rates were derived from child interview data. Results indicated that four symptoms served as efficient inclusion (high PPPs) and exclusion (high NPPs) criteria for depressive disorders: feeling unloved, anhedonia, excessive guilt, and depressed mood. Symptoms describing worries served as the most efficient inclusion criteria for anxiety disorders, especially worries about future events and competence in academics. However, based on the PPP values, the anxiety symptoms were more efficient predictors of the depression diagnosis than the anxiety diagnosis. This heuristic procedure is presented as a model to establish the most efficient inclusion and exclusion criteria for the diagnoses of childhood internalizing disorders.
NASP Members Log in
to download article.