Sensitivity and Specificity of the SRQ-20 and the EPDS in Diagnosing Major Depression Ante- and Postnatally in a South African Birth Cohort Study

Abstract

We assessed the sensitivity and specificity of the Self-Reporting Questionnaire (SRQ-20) and the Edinburgh Postnatal Depression Scale (EPDS), against the major depression module of the Mini International Neuropsychiatric Interview (MINI). Data were utilised from antenatal (n = 296) and postnatal participants (n = 366) in the Drakenstein Child Health Study (DCHS), a multidisciplinary birth cohort investigating the determinants of child health. Mothers were interviewed using the SRQ-20, the EPDS and the MINI. Receiver operating characteristic (ROC) curve analysis was performed in order to establish optimal cut-off scores. Current major depressive episode was diagnosed in 5% of antenatal and 4% of postnatal participants. At widely used threshold scores, the sensitivity of the tools ranged from 50 to 67%, although the positive predictive values at these scores were much lower (14–25%). Both tools perform acceptably in identifying mothers at risk for major depression perinatally. In identifying appropriate cut-off scores, optimising specificity and maintaining at least 30% sensitivity seems to be a feasible approach given the resourcing of South African mental health services.