Evaluation of Repeated Web-Based Screening for Predicting Postpartum Depression: Prospective Cohort Study

Area of Science:

• Reproductive Health
• Mental Health
• Digital Health Technologies

Background:

• Postpartum depression (PPD) negatively impacts maternal-infant bonding and infant development.
• Universal screening for PPD is recommended, but optimal timing during pregnancy remains undefined.
• Web-based technologies offer opportunities for early detection of high-risk pregnancies.

Purpose of the Study:

• To stratify PPD risk and determine optimal prenatal screening times.
• To develop a screening algorithm using a web-based tool for electronic patient-reported outcomes (ePROs).

Main Methods:

• 214 women completed the Edinburgh Postnatal Depression Scale (EPDS) via a web platform 5 times during pregnancy and 3 times postpartum.
• Area Under the Curve (AUC), sensitivity, specificity, and predictive values were calculated for prenatal assessments.
• Multivariate logistic regression identified predictors like age, education, parity, relationship quality, and anxiety to enhance accuracy.

Main Results:

• Screening in the second (AUC=0.85) and third (AUC=0.75) trimesters showed reasonable to good predictive accuracy for PPD.
• Multivariate logistic regression yielded excellent AUC (0.93) at 3 months and good AUC (0.87) at 6 months postpartum.
• The optimal predictive accuracy was observed for screening between 24-28 gestational weeks.

Conclusions:

• Screening for PPD between 24-28 gestational weeks is beneficial for identifying at-risk women.
• Integrating personal risk factors significantly improves predictive discriminatory power, particularly for 3-month postpartum outcomes.
• Web-based tools are feasible, efficient, and cost-effective for early detection of high-risk pregnancies, improving long-term maternal and infant outcomes.