Accuracy of the Edinburgh Postnatal Depression Scale in screening for major depressive disorder and other psychiatric disorders in women towards the end of their puerperium.

Abstract

To assess the accuracy of the Edinburgh Postnatal Depression Scale (EPDS) in screening for severe depression and other mental disorders in women at the end of puerperium. We administered the Czech version of the EPDS to assess depressive symptoms and the Mini International Neuropsychiatric Interview to determine psychiatric dia-gnoses in 243 women at the end of their puerperium. Then, we determined the frequencies of severe depressive disorder and other psychiatric disorders in our cohort. Furthermore, we assessed the sensitivity, specificity, positive predictive value, negative predictive value, and other dia-gnostic variables for the presence of severe depression and other psychiatric disorders for different threshold scores on EPDS. We evaluated the detection potential of EPDS for detecting monitored mental disorders by using the receiver operating characteristic curve analysis and determining the area under the curve. Severe depressive disorder was present in 2.5% (95% CI: 1.1-5.3%) of women. Any monitored mental disorder was present in 13.6% (95% CI: 9.8-18.5%). The best sensitivity/specificity ratio for detecting major depressive disorder was found for the EPDS threshold score 11; sensitivity was 83% (95% CI: 35-99%) and specificity was 79% (95% CI: 74-84%). The EPDS 11 then achieved a sensitivity of 76% (95% CI: 58-89%) and specificity of 82% (95% CI: 76-87%) for the detection of any mental disorder of interest. Our results showed that the Czech version of EPDS has good internal consistency, and the EPDS score 11 achieves the best combination of sensitivity and specificity values for detecting major depressive disorder. Screening with EPDS in women at the end of puerperium can detect psychiatric disorders other than severe major depression.