Diagnostic errors of primary care screens for depression and panic disorder

As the health care reimbursement system has changed, brief screens for detecting mental disorders in primary care have been developed. These efforts have faced the formidable task of identifying patients with mental disorders, while at the same time minimizing the number of misclassified cases. Here we consider the balance between sensitivity and positive predictive value. Primary care patients with false positive and false negative results on screens for depression and panic disorder are compared with regard to comorbidity and functional impairment.

This was a cross-sectional psychometric study. The study sample included 1001 primary care patients from the Department of Internal Medicine at Kaiser Permanente in Oakland, California. The Symptom-Driven Diagnostic System for Primary Care (SDDS-PC) screens and Sheehan Disability Scale were completed by the subjects. The SDDS-PC diagnostic interviews were administered to all subjects.

Patients with false positive results on the panic disorder screen did not differ from patients with false negatives results with regard to rates of other psychiatric disorders, functional impairment, or mental health service utilization. In contrast, patients with false negative depression screen results had significantly more psychiatric disorders and functional impairment than those with false positive depression results.

A substantial number of patients with either false positive or false negative screen results met diagnostic criteria for other mental disorders. Given the nominal burden of follow-up assessments for patients with positive screens, these data suggest that erring on the side of sensitivity may have been preferable when algorithms for these screens were selected.