Postpartum depression (PPD) is one of the most common complications of childbearing, estimated to affect between 10% and 15% of mothers worldwide. It is a leading cause of maternal perinatal mortality, and has a negative association with infant cognitive development, language development, and behaviors. The current screening routine is broadly based on identifying symptoms using self-reported questionnaires such as the Edinburgh Postnatal Depression Scale (EPDS).
We harness electronic health record (EHR) data of women who have given birth and identify PPD based on diagnosis codes, drug prescriptions, and non-pharmacological treatments. Using machine-learning techniques, we develop prediction models to identify women at risk. These models may improve current screening tools and facilitate early intervention.