Machine Learning-Based Prediction of Maternal Morbidity across Heterogeneous Populations in the United States using Sequential Modeling of the All of Us Dataset
- 1UC Berkeley
Abstract
In this work, we demonstrate the unprecedented value of NIHs "All of Us Research Program" (AoURP) dataset in studying maternal morbidity and building predictive machine learning (ML) models across heterogeneous populations in the United States. We developed robust and data-driven preprocessing pipelines to curate a longitudinal, multi-site, multimodal, and demographically diverse pregnancy dataset (20,253 subjects; 27,525 pregnancy episodes) from AoURP data, using electronic health records (EHR) (Conditions, Labs, Measurements) and survey responses (Social Determinant of Health (SDoH)), focusing on 7 crucial maternal health adverse outcomes. After characterizing data quality, missingness, and heterogeneity, we performed statistical correlation analysis to identify risk factors. We subsequently developed XGBoost and sequential LSTM models to predict the adverse outcomes, reaching state-of-the-art performance for multiple outcomes. We conducted model interpretability post-hoc analysis to understand success points and fairness analysis to evaluate implications for socio-economic disparities. Four practicing physicians reviewed the set of statistically significant and ML model identified features to assess their clinical validity and novelty. Most features identified through either statistical correlations or ML feature importance analysis aligned with known clinical risk factors. Several features were identified that the ML models used but that are not currently used in clinical practice and may merit further clinical investigation. Fairness analysis revealed certain associations with SDoH and age highlight areas that warrant continued monitoring. Overall, we demonstrate that meaningful populational level patterns can be extracted, and high-performing machine learning models can be trained on this longitudinal, diverse, multi-site dataset. Important risk features, particularly novel ones identified, if validated, could inform new strategies for maternal care or enable development and validation of outcome-specific, clinically deployable ML models.
Related
- PRInter: an inductive multimodal heterogeneous network embedding framework with contrastive learning for lncRNA-protein interaction prediction
- Hyperbolic heterogeneous graph learning for adverse drug reaction prediction, candidate prioritization, and pathway-based interpretation in pharmacovigilance
- Heterogeneous and dynamic lung cancer mortality among immigrants relative to native-born populations in France, 2000–2021
- Personalized prediction of sequential therapeutic responses in HR+/HER2-low breast cancer using patient-derived organoids
- A clinical prediction nomogram for hemorrhoid risk in health screening populations in Beijing, China: a retrospective cohort study of 35,256 adults
BibTeX
@misc{zhuang2026machine,
title = {{Machine Learning-Based Prediction of Maternal Morbidity across Heterogeneous Populations in the United States using Sequential Modeling of the All of Us Dataset}},
author = {H. Zhuang and A. Zakama and K. Heller and S. Faulkner and B. Gollub and N. Young-Lin and I. Y. Chen and M. Asiedu},
howpublished = {medRxiv},
year = {2026},
month = aug,
doi = {10.64898/2026.08.25.26360552},
url = {https://www.medrxiv.org/content/10.64898/2026.08.25.26360552v1},
note = {CC BY},
}
RIS
TY - UNPB
TI - Machine Learning-Based Prediction of Maternal Morbidity across Heterogeneous Populations in the United States using Sequential Modeling of the All of Us Dataset
AU - H. Zhuang
AU - A. Zakama
AU - K. Heller
AU - S. Faulkner
AU - B. Gollub
AU - N. Young-Lin
AU - I. Y. Chen
AU - M. Asiedu
PY - 2026
DA - 2026/08/25
PB - medRxiv
DO - 10.64898/2026.08.25.26360552
UR - https://www.medrxiv.org/content/10.64898/2026.08.25.26360552v1
AB - In this work, we demonstrate the unprecedented value of NIHs "All of Us Research Program" (AoURP) dataset in studying maternal morbidity and building predictive machine learning (ML) models across heterogeneous populations in the United States. We developed robust and data-driven preprocessing pipelines to curate a longitudinal, multi-site, multimodal, and demographically diverse pregnancy dataset (20,253 subjects; 27,525 pregnancy episodes) from AoURP data, using electronic health records (EHR) (Conditions, Labs, Measurements) and survey responses (Social Determinant of Health (SDoH)), focusing on 7 crucial maternal health adverse outcomes. After characterizing data quality, missingness, and heterogeneity, we performed statistical correlation analysis to identify risk factors. We subsequently developed XGBoost and sequential LSTM models to predict the adverse outcomes, reaching state-of-the-art performance for multiple outcomes. We conducted model interpretability post-hoc analysis to understand success points and fairness analysis to evaluate implications for socio-economic disparities. Four practicing physicians reviewed the set of statistically significant and ML model identified features to assess their clinical validity and novelty. Most features identified through either statistical correlations or ML feature importance analysis aligned with known clinical risk factors. Several features were identified that the ML models used but that are not currently used in clinical practice and may merit further clinical investigation. Fairness analysis revealed certain associations with SDoH and age highlight areas that warrant continued monitoring. Overall, we demonstrate that meaningful populational level patterns can be extracted, and high-performing machine learning models can be trained on this longitudinal, diverse, multi-site dataset. Important risk features, particularly novel ones identified, if validated, could inform new strategies for maternal care or enable development and validation of outcome-specific, clinically deployable ML models.
ER -