Limits of Machine Learning for Predicting Stillbirth at Term: A Population-Based Cohort of 302,192 Pregnancies in Northern Italy

Authors

• M. Benuzzi, University of Modena and Reggio Emilia
• T. Kawakita, Macon & Joan Brock Virginia Health Sciences at Old Dominion UniversityFollow
• F. Monari, University of Modena and Reggio Emilia
• B. Melis, University of Modena and Reggio Emilia
• G. Guariglia, University of Modena and Reggio Emilia
• A. La Marca, University of Modena and Reggio Emilia
• I. Neri, University of Modena and Reggio Emilia
• G. Saade, Macon & Joan Brock Virginia Health Sciences at Old Dominion UniversityFollow
• F. Facchinetti, University of Modena and Reggio Emilia
• I. El Moudden, Old Dominion UniversityFollow

ORCID

0000-0003-2438-1699 (Kawakita), 0000-0001-7925-5302 (Saade), 0000-0001-7702-2564 (El Moudden)

Document Type

Article

Publication Date

2026

DOI

10.1016/j.ajogmf.2026.102148

Publication Title

American Journal of Obstetrics & Gynecology MFM

Volume

Advance online publication

Pages

40 pp.

Abstract

Background

Stillbirth at term is a major public-health concern. In high-income settings, the most commonly identifiable causes are unrecognised fetal growth restriction (about 14%) and placental abruption (about 12%), but 20 to 40% of term stillbirths remain unexplained. Whether modern machine learning can substantively improve prediction over established clinical risk-factor models is unresolved.

Objective

To evaluate the predictive performance of three machine-learning models, elastic-net logistic regression, random forest, and extreme gradient boosting, for stillbirth at term in a population-based cohort, and to quantify the achievable ceiling using static registry variables.

Study Design

Secondary analysis of 302,192 term pregnancies (322 stillbirths; rate 1.07 per 1,000), recorded in the Emilia-Romagna “Certificato di Assistenza al Parto” birth registry between 2014–2023. Three machine-learning models were trained on a stratified 80% training set with class rebalancing by Random Over-Sampling Examples applied within each cross-validation fold. Logistic-regression predicted probabilities were rescaled to natural prevalence by intercept-only Platt-style recalibration on a held-out training slice. Performance was evaluated on the held-out 20% test set (n=60,438; 66 stillbirths) using area under the receiver operating characteristic curve (AUC-ROC), area under the precision-recall curve (AUC-PR), sensitivity, specificity, predictive values (Wilson 95% CIs), calibration intercept and slope, observed:expected (O:E) ratio, and decision curve analysis. Sensitivity analyses included a head-to-head comparison with the previously published Salerno (2023) logistic regression specification, complete-case and missing-as-zero analyses, feature ablation, refitting without class rebalancing, bootstrap optimism correction, and temporal validation (training 2014-2021; test 2022-2023).

Results

On the test set, elastic-net logistic regression achieved AUC 0.658 (0.589-0.726), XGBoost AUC 0.617 (0.544-0.690), and random forest AUC 0.534 (0.470-0.597). All three fell below the AUC of 0.70 threshold for clinical screening. Logistic regression significantly outperformed random forest (DeLong p=0.001). Positive predictive values at Youden-optimal thresholds were within Wilson-interval distance of the population baseline (0.106%). Decision curve analysis showed no model produced net benefit superior to a treat-all strategy at clinically plausible thresholds. After intercept recalibration, the logistic regression was well calibrated in the large (observed:expected ratio 1.12, 95% CI 0.88 to 1.43). Gestational age, particularly early-term delivery, was the most influential predictor across models. Temporal validation degraded all three AUCs to 0.57 to 0.58. The previously published Salerno (2023) logistic regression specification achieved AUC 0.637 on the same test set, only 0.02 below the elastic-net logistic regression.

Conclusions

In a large, decade-spanning population cohort, three machine-learning models achieved similarly limited discrimination for term stillbirth (AUC 0.53 to 0.66), with the simpler regularised regression significantly outperforming the random forest. Static registry variables, even those captured by a high-quality regional surveillance system, do not carry sufficient signal for clinically useful term-stillbirth prediction. The empirical ceiling defined by this study should motivate a shift toward dynamic, physiology-based surveillance for future prediction research.

Rights

© 2026 The Authors.

This is an open access article under the Creative Commons Attribution 4.0 International (CC BY 4.0) License.

Original Publication Citation

Benuzzi, M., Kawakita, T., Monari, F., Melis, B., Guariglia, G., La Marca, A., Neri, I., Saade, G., Facchinetti, F., & El Moudden, I. (2026). Limits of machine learning for predicting stillbirth at term: A population-based cohort of 302,192 pregnancies in northern Italy. American Journal of Obstetrics & Gynecology MFM. Advance online publication. https://doi.org/10.1016/j.ajogmf.2026.102148

Repository Citation

Benuzzi, M., Kawakita, T., Monari, F., Melis, B., Guariglia, G., La Marca, A., Neri, I., Saade, G., Facchinetti, F., & El Moudden, I. (2026). Limits of machine learning for predicting stillbirth at term: A population-based cohort of 302,192 pregnancies in northern Italy. American Journal of Obstetrics & Gynecology MFM. Advance online publication. https://doi.org/10.1016/j.ajogmf.2026.102148