Machine learning-based predictive model for postpartum post-traumatic stress disorder: A prospective cohort study

Area of Science:

• Perinatal mental health research
• Computational psychiatry
• Clinical informatics

Background:

• Postpartum Post-Traumatic Stress Disorder (PTSD) presents a significant public health challenge for mothers and infants.
• Early identification of women at high risk for postpartum PTSD is crucial for mitigating adverse outcomes.
• This study focused on developing and validating a machine learning (ML) model for early postpartum PTSD risk prediction.

Purpose of the Study:

• To develop and validate a machine learning (ML)-based model for predicting the risk of Postpartum Post-Traumatic Stress Disorder (PTSD).
• To enable early identification of high-risk women for timely and targeted interventions.
• To improve maternal mental health care through the application of predictive analytics.

Main Methods:

• A prospective cohort study involving 900 women for model development and 300 for validation.
• Data collected included sociodemographic, clinical, adverse childhood experiences, and biochemical factors at 3 days postpartum.
• Five ML models were evaluated, with performance assessed using discrimination, calibration, and Area Under the ROC Curve (AUC).

Main Results:

• The Logistic Regression (LR) model demonstrated the highest predictive performance with an AUC of 0.850 (95% CI: 0.776-0.923).
• The LR model achieved a Brier score of 0.069, sensitivity of 0.844, and specificity of 0.724.
• A web-based risk calculator, utilizing 8 key predictors, was developed for clinical application.

Conclusions:

• Machine learning models are effective in predicting postpartum PTSD risk.
• The developed web-based risk calculator facilitates early identification of at-risk mothers.
• Further validation in diverse cohorts is recommended to enhance the generalizability of ML tools in maternal mental health.