Screening for postnatal depression in routine primary care: properties of the Edinburgh Postnatal Depression Scale in an Australian sample

OBJECTIVES: First, to explore the utility of the Edinburgh Postnatal Depression Scale (EPDS) in routine primary care through a large community screening program. Next, to compare administration of a second EPDS versus the Beck Depression Inventory (BDI) in identifying postnatal depression in the prescreened population. METHOD: Screening with the EPDS through Maternal and Child Health Centres at 4 months post-partum. Women scoring > or = 12 were assessed against DSM-IV criteria and completed a BDI and a second EPDS. These data were subjected to receiver operating characteristic (ROC) analyses. RESULTS: Of 4148 screened, 533 (12.8%) scored > or = 12. Of these, 344 were assessed against DSM-IV criteria: 193 (56%) - major depressive disorder; 67 (20%) - other diagnoses that incorporated depression. Positive predictive value at screening was therefore 76%. Another 45 (13%) had non-depressive disorders and 39 (11%) were psychiatric non-cases. The BDI was the better diagnostic instrument in the prescreened population, having a significantly higher efficiency as quantified by ROC curve analysis, though the absolute difference in efficiency was small (approximately 6%). CONCLUSIONS: Screening with the EPDS integrated well into routine primary care. Two-step screening offers one way of achieving acceptable balances of operational simplicity and diagnostic accuracy.