Screening for anxiety in an epidemiological sample: predictive accuracy of questionnaires.

The study examined the predictive accuracy of selected questionnaires when screening for anxiety in a representative epidemiological sample of young female adults (N = 1877). All participants were diagnosed using a structured diagnostic interview. Anxiety questionnaires included global as well as specific measures (Beck Anxiety Inventory (BAI), Symptom Checklist, Anxiety Sensitivity Index (ASI), Fear Questionnaire (FQ), Mobility Inventory (MI)). Sensitivity, specificity, positive and negative predictive power were computed for two screening decisions: (1) identifying any anxiety disorder or (2) identifying a specific anxiety disorder (agoraphobia) within the total sample and the clinical subsample. Due to naturalistic (low) base rates in epidemiological samples, diagnostic indices were lower than those previously reported. However, questionnaire data proved useful when a specific disorder was targeted (agoraphobia) and specific symptoms were operationalized.