The screening accuracy of the Edinburgh Postnatal Depression Scale (EPDS) to detect perinatal depression with and without the self-harm item in pregnant and postpartum women.

Background: This study aims to examine whether the Edinburgh Postnatal Depression Scale (EPDS), excluding the self-harm item (EPDS-9), performs as effectively as the full EPDS in identifying depression among perinatal women.

Methods: A total of 3571 pregnant women and 3850 postpartum women participated in this observational study. Participants who scored ≥ 9 on the EPDS underwent further diagnostic evaluations by a clinical psychologist and/or psychiatrist.

Results: The EPDS-9 and full EPDS demonstrated a near-perfect correlation in both the antepartum (r = 0.996) and postpartum (r = 0.998) cohorts. EPDS-9 showed exceptional precision in identifying depression as screened by the full EPDS at cutoff points ranging 9-14, with areas under the curve ≥0.998. The sensitivity of EPDS-9 and full EPDS to detect depression that requires psychotropic medications was poor. The highest accuracy for both versions was at a cutoff score of 9: sensitivity of 0.579 for the full EPDS and 0.526 for the EPDS-9. At the cutoff point of 9, EPDS-9 performed adequately in predicting the response of the participants to the self-harm item.

Conclusion: The EPDS-9 represents a solid and effective replacement for the full EPDS in clinical settings. If the presence of suicidal thoughts needs to be assessed, specialized scales should be used.