Diagnosis of bipolar II disorder: a comparison of structured versus semistructured interviews

Background: Reliability of bipolar II (BPII) disorder diagnosis is still a problem. Recent studies have shown that semistructured interviews by clinicians are better than structured interviews by nonclinicians for BPII diagnosis. The aim of the study was to find the degree of agreement in the diagnosis of BPII between the Structured Clinical Interview for DSM-IV (SCID) and a semistructured interview based on DSM-IV criteria done by an expert clinician. Methods: One hundred eleven remitted major depressive episode (MDE) outpatients were interviewed first with the SCID and soon after that with a semistructured interview following DSM-IV criteria (based on clinical evaluation). Bipolar I (BPI) patients were excluded. Results: By the SCID, 24 patients were diagnosed BPII (21.6%) and 30 were diagnosed BPI (27.0%). By the semistructured interview, 68 patients were diagnosed BPII (61.2% of the entire sample) and none BPI. Agreement between the SCID BPII diagnosis and the semistructured interview BPII diagnosis was 51.3% (meaning one in two missed). Sensitivity and specificity of the SCID BPII diagnosis for the semistructured BPII diagnosis were 29.4% and 90.7%, respectively. Overactivity (increased goal-directed activity) was the most common hypomanic symptom. In the group with overactivity (n = 76), a semistructured interview BPII diagnosis was present in 77.6%, while a SCID BPII diagnosis was present in only 22.3%. Sensitivity and specificity of overactivity for BPII diagnosis were 86.7% and 60.4%, respectively, while elevated mood had sensitivity of 60.2% and specificity of 86.0%. Conclusions: Findings support a diagnosis of BPII based on a semistructured interview by an expert clinician versus a fully structured interview. Overactivity priority level for the diagnosis of hypomania is supported by the present findings. (C) 2003 Elsevier Inc. All rights reserved.