Women with bipolar disorder (BD) are at extremely high risk for postpartum psychiatric illness, specifically postpartum psychosis.  While the American College of Obstetrics and Gynecology now recommends that all women be screened for depression during pregnancy and the postpartum period, we have little information regarding the screening tools that are best suited for the identification of women with bipolar disorder.  Many sites now use the Edinburgh Postnatal Depression Scale for postpartum screening.  This instrument is fairly good for identifying women with unipolar depression, but how well does it do in terms of identifying women with bipolar disorder?

Researchers recently evaluated the Mood Disorders Questionnaire (MDQ) and the EPDS for the identification of postpartum women with bipolar disorder.    Postpartum women (N = 1,279) were screened with the EPDS and MDQ at 4-6 weeks after delivery. Women who had positive screens were evaluated with the Structured Clinical Interview for DSM-IV (SCID) to establish psychiatric diagnosis.

12% of the women screened positively on the EPDS and/or the MDQ (n = 155).  SCID diagnostic interviews were completed in 93 (60%) of the mothers with positive screens.   Bipolar disorder was the primary diagnosis in 34 (37%) of the women with positive screens.

Women with BD screened positive on the EPDS and/or MDQ as follows:   EPDS+/MDQ+ (n = 14), EPDS+/MDQ- (n = 17), EPDS-/MDQ+ (n =3).   The MDQ identified 50% (17/34) of the women with BD; however, the MDQ picked up 23/34 or 68% of the cases when the MDQ question regarding the mother’s perception of herself as being impaired was excluded.

This study suggests that if we use the EPDS alone, we would miss about 10% of women with bipolar disorder (3 out of 34); that’s a 90% detection rate, which is respectable for a screening tool.   Where the EPDS seems to fail us is in terms of its diagnostic accuracy or specificity.  If we just used the EPDS to screen postpartum women, we would find that about one-third of the women who screen positive on the EPDS actually have bipolar disorder (as opposed to unipolar depression).

The EPDS was not designed to be a diagnostic tool, and when it is used in this manner, things can go awry.  The EPDS is a very good tool for identifying women with psychiatric illness, but to precisely determine the nature of that illness, the woman needs a psychiatric evaluation.  Adding the MDQ improves, to some degree, our ability to distinguish bipolar disorder from major depression; however, using two screening tools cannot be used as a substitute for a thorough psychiatric evaluation.  In order to provide the appropriate treatment, we need first to make an accurate diagnosis.  For women with bipolar disorder, misdiagnosis can delay treatment and, at the worst, may actually lead to interventions which exacerbate the clinical picture.

As we move forward with screening, we need to make sure we know how to appropriately refer women who are identified using these screening tools.  The ACOG guidelines urge “clinical staff in obstetrics and gynecology practices should be prepared to initiate medical therapy, refer patients to appropriate behavioral health resources when indicated, or both.” I suspect, however, that many obstetric providers may feel uncomfortable assuming the psychiatric care of this patient population.  While screening is important, we must also make sure we tend to the ACOG recommendations regarding appropriate follow-up and treatment.   Because the stigma continues to be significant with regard to mental health issues in mothers and mothers-to-be and because there are concerns regarding the use of medication in pregnant and nursing women, we must make sure that after screening, we help women to access appropriate resources and treaters who are familiar with treating women during pregnancy and the postpartum period.

Ruta Nonacs, MD PhD

Clark CT, Sit DK, Driscoll K, et al.  DOES SCREENING WITH THE MDQ AND EPDS IMPROVE IDENTIFICATION OF BIPOLAR DISORDER IN AN OBSTETRICAL SAMPLE? Depress Anxiety. 2015 Jul;32(7):518-26.