Maya Singh, Meredith Spada, Michelle Wright, Samantha Rodriguez, Karen Jakubowski, Heather Joseph, Lora Johnson, Lauren Kullen, Cheryl Siverling, Kaleigh Gustafson, Jennifer Chaney, Lara Lemon, Tamar Krishnamurti, Priya Gopalan
Delivery-based bedside screening identified individuals at elevated risk for postpartum depressive symptoms, while EHR-derived variables provided scalable risk stratification. Integrating EHR-based indicators alongside psychosocial screening may support more proactive and individualized perinatal mental health care.
OBJECTIVE: To evaluate a delivery-based bedside screening workflow for identifying risk of postpartum depressive symptoms within one year postpartum and compare its performance with a parsimonious electronic health record (EHR)-derived approach.
METHODS: This retrospective cohort study included individuals screened during delivery hospitalization at two hospitals within a single integrated health system between November 2023 and November 2024. A structured bedside screening tool incorporating psychiatric, psychosocial, and obstetric risk factors was administered at admission. The primary outcome was elevated postpartum depressive symptoms, defined as Patient Health Questionnaire-9 (PHQ-9) score ≥ 10 within one year postpartum. Associations between screening variables and postpartum PHQ-9 scores were evaluated using nonparametric analyses. A parsimonious multivariable logistic regression model incorporating routinely available EHR-derived variables was also evaluated.
RESULTS: Among 575 individuals with postpartum PHQ-9 data, 67 (11.7%) had PHQ-9 scores ≥10 postpartum. In the analytic cohort (n = 512), individuals classified as high risk at delivery had higher postpartum PHQ-9 scores than those not classified as high risk (median 3 vs 1, p < 0.001). Low social support, intimate partner violence, prior postpartum depression, and PHQ-2 ≥ 3 demonstrated strong univariate associations with postpartum symptom severity. A parsimonious EHR-based model including prenatal PHQ-9 score and insurance status demonstrated higher discrimination for PHQ-9 ≥ 10 than the bedside high-risk designation (AUC 0.74 vs 0.61).
CONCLUSIONS: Delivery-based bedside screening identified individuals at elevated risk for postpartum depressive symptoms, while EHR-derived variables provided scalable risk stratification. Integrating EHR-based indicators alongside psychosocial screening may support more proactive and individualized perinatal mental health care.