Related Experiment Video
Updated: Jan 11, 2026
07:31
Implementation of a Real-Time Psychosis Risk Detection and Alerting System Based on Electronic Health Records using CogStack
Published on: May 15, 2020
7.5K
Identifying determinants and predicting cesarean section delivery among Bangladeshi women using machine learning:
Shamsuz Zoha1, Shahin Alam1, Isteaq Kabir Sifat1
1Department of Statistics, Hajee Mohammad Danesh Science and Technology University, Dinajpur, Bangladesh.
PLOS Global Public Health
|November 19, 2025
Summary
Rising Cesarean section (C-section) rates are concerning. Machine learning models effectively identified key risk factors, enabling better prediction and intervention strategies for maternal healthcare in Bangladesh.
Area of Science:
- Public Health
- Medical Informatics
- Reproductive Health
Background:
- Global Cesarean section (C-section) rates are increasing, necessitating research into contributing factors and predictive modeling.
- Understanding C-section determinants is crucial for developing targeted interventions and informing health policies.
- Machine learning (ML) offers powerful tools for analyzing complex health data and predicting delivery outcomes.
Purpose of the Study:
- To identify key determinants influencing C-section delivery in Bangladesh.
- To develop and evaluate an ML-based predictive model for C-section risk.
- To leverage findings for improved maternal healthcare planning and potential reduction of unnecessary C-sections.
Main Methods:
- Analysis of 2,490 records from the Bangladesh Demographic and Health Survey (BDHS) 2022.
- Utilized Recursive Feature Elimination (RFE), Boruta-based selection (BFS), and Random Forest (RF) for feature selection.
- Employed six ML algorithms (LR, SVM, KNN, DT, RF, XGB) for prediction, evaluated using accuracy, precision, recall, F1-score, AUC, and ROC analysis. SHAP values were used for feature interpretation.
Main Results:
- C-section prevalence was 45.6%. Ten significant determinants were identified: place of delivery, baby weight, maternal BMI, birth interval, age at first birth, partner's education, maternal age, wealth status, ANC visits, and maternal education.
- The Random Forest (RF) model demonstrated the highest performance with 81.79% accuracy and an AUC of 0.871.
- SHAP analysis indicated place of delivery, baby weight, maternal BMI, and birth interval as the most influential predictors.
Conclusions:
- Socio-demographic and healthcare factors significantly influence C-section delivery rates in Bangladesh.
- Machine learning models, particularly RF, are effective in identifying high-risk individuals for C-section.
- The study provides a foundation for data-driven strategies to reduce unnecessary C-sections and enhance maternal care planning.
