Researchers at Mass General Brigham have developed a machine learning model that predicts postpartum depression (PPD) risk using routine clinical and demographic data available at delivery.
Published in the American Journal of Psychiatry, the study highlights the model’s potential to identify at-risk patients earlier than traditional postpartum screenings.
PPD affects up to 15% of new parents, often going undetected until the standard 6-to-8 week postpartum visit. The new model analyzes electronic health record (EHR) data—including demographics, medical history, and visit records—from over 29,000 patients across multiple hospitals.
It accurately ruled out PPD in 90% of cases and identified nearly 30% of high-risk individuals who developed PPD within six months after childbirth.
Importantly, the model’s performance was consistent across different races, ethnicities, and ages, and it worked well even among patients without prior psychiatric diagnoses. Incorporating prenatal Edinburgh Postnatal Depression Scale scores further enhanced its predictive power.
The research team is now testing the model prospectively and collaborating with clinicians and patients to integrate this tool into routine care, aiming to provide earlier mental health support and improve outcomes for new parents.
YOU MAY ALSO LIKE:
