Screening for bipolar disorder among outpatients with substance use disorders

Abstract Background Comorbidity of bipolar disorder and alcohol or substance abuse/dependence is frequent and has marked negative consequences on the course of the illness and treatment compliance. The objective of this study was to compare the validity of two short instruments aimed at screening bipolar disorders among patients treated for substance use disorders. Methods The Mood Disorder Questionnaire (MDQ) and the Hypomania Checklist-32 (HCL-32) were tested with reference to the mood section of the Structured Clinical Interview for DSM-IV axis I disorders (SCID) in 152 patients, recruited in two outpatient clinics providing specialized treatment for alcohol and opiate dependence. Results According to the SCID, 33 patients (21.7%) had a diagnosis within the bipolar spectrum (two bipolar I, 21 bipolar II and 10 bipolar not otherwise specified). The HCL-32 was more sensitive (90.9% vs. 66.7%) and the MDQ more specific (38.7% vs. 77.3%) for the whole sample. The MDQ displayed higher sensitivity and specificity in patients treated for alcohol than for opiate dependence, whereas the HCL-32 was highly sensitive but poorly specific in both samples. Both instruments had a positive predictive value under 50%. Conclusions Caution is needed when using the MDQ and HCL-32 in patients treated for substance use disorders.