Immediate postpartum depressive symptoms and future psychiatric diagnoses: A predictive analysis using the EPDS.

Abstract

Background: The implementation of universal screening to identify women at risk of postnatal mental disorders has been recommended. Understanding whether depressive symptoms immediately postpartum, as detected by screening tools, predict later psychiatric diagnoses is crucial for guiding timely care and optimizing healthcare resources.

Methods: All participants completed the Edinburgh Postnatal Depression Scale (EPDS) 2-5 days postpartum. Women with scores ≥ 10 were invited to the follow-up study, where psychiatric diagnoses at six weeks were assessed using the Mini-International Neuropsychiatric Interview 5 (MINI). Sensitivity, specificity, positive (PPV), and negative (NPV) predictive values were calculated. ROC analyses assessed predictive accuracy.

Results: A total of 403 women were included. EPDS scores of ≥ 16 for a major depressive episode (AUC = 0.78, sensitivity 0.71, specificity 0.84, PPV 0.14, NPV 0.99 within this high-risk sample) and ≥ 14 for any observed mental disorder (AUC = 0.69, sensitivity 0.65, specificity 0.73, PPV 0.24, NPV 0.94) were identified as optimal.

Conclusions: In women with elevated depressive symptoms shortly after delivery, the EPDS showed high NPV, supporting its usefulness for ruling out severe psychopathology at six weeks postpartum. However, the low PPV indicates that elevated scores often reflect transient postpartum distress rather than a confirmed disorder. Routine immediate-postpartum EPDS screening should therefore be approached with caution. Findings support a tiered approach-for example, brief initial inquiries followed by the EPDS and diagnostic follow-up for higher scores-and the development of screening tools better adapted to the immediate postpartum period.