Optimization of the Mood Disorder Questionnaire in identification of perinatal bipolar disorder

BACKGROUND: The recognition of bipolar disorder during the perina-tal period is often challenging because birthing people most commonly present in a depressive episode. The phenotypic expression of episodes of bipolar depression is difficult to differentiate from major depressive disor-der and can lead to misdiagnosis and inappropriate treatment. The Mood Disorder Questionnaire is a readily available screening tool for bipolar dis-order that has been validated in previous studies for use in the general and perinatal populations. However, the discriminatory capacity of the Mood Disorder Questionnaire for perinatal people who screen positive for depression in nonpsychiatric settings is still unclear. OBJECTIVE: This study aimed to evaluate the discriminatory capacity of the Mood Disorder Questionnaire to identify bipolar disorder in perinatal people who screen positive for depression on the Patient Health Question-naire-9. STUDY DESIGN: This retrospective cohort study included individuals enrolled in the Collaborative Care Model for Perinatal Depression Support Services, a collaborative care program for perinatal mental health services implemented in a quaternary care setting, from January 2017 to April 2021. All individuals completed the Mood Disorder Questionnaire and psy-chiatric evaluation by a licensed clinical social worker. Clinical and socio-demographic characteristics were compared between those with and without a clinical diagnosis of bipolar disorder using bivariable analyses. The discriminatory capacity and test characteristics of the Mood Disorder Questionnaire were assessed at each score cutoff using the gold standard of a psychiatric clinical evaluation for comparison.RESULTS: From January 2017 to April 2021, 1510 birthing people were enrolled in the Collaborative Care Model for Perinatal Depression Support Services and included in this study. Among this group, 62 (4.1%) were diagnosed with bipolar disorder by psychiatric clinical evaluation using the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition, diagnostic criteria. A score of & GE;7 on question 1 is often used in the general population to identify bipolar disorder, which has a 60% sensi- tivity and 88% specificity in our perinatal sample with an area under the receiver operating characteristic curve of 0.74 (95% confidence interval, 0.72-0.76). Lowering the threshold to & GE;4 improves sensitivity to 81% and the discriminatory capacity to an area under the receiver operating characteristic curve of 0.75 (95% confidence interval, 0.70-0.80), at the expense of a reduction in specificity to 69%.CONCLUSION: The administration of the Mood Disorder Questionnaire in the perinatal period can help to identify which individuals who have screened positive for depression on the Patient Health Questionnaire-9 are at risk of a bipolar or unipolar disorder. In this context, lowering the Mood Disorder Questionnaire score threshold from that used in the non- perinatal population down to 4 improves test characteristics and reduces the risk of a missed diagnosis of bipolar disorder.