COIOSDeep risk insights through multi-disciplinary research across the globe
A view of the ocean from a hill
Item refines Peer-reviewed Added

Machine learning on routine records predicts three pregnancy complications by 12 weeks

In 442,760 Kaiser Permanente pregnancies, XGBoost models using only routine data predicted gestational diabetes, gestational hypertension and pre-eclampsia at 12 weeks with AUCs of 0.74 to 0.80, but birth outcomes were harder to predict.

The study tested machine learning on routine health records in 442,760 Kaiser Permanente Northern California pregnancies, 2011 to 2021. Separate XGBoost models predicted seven complications at six-week points, using only data recorded before each point. At 12 weeks, AUCs were 0.800 for gestational diabetes, 0.789 for gestational hypertension and 0.740 for pre-eclampsia. At 24 weeks, AUCs for preterm birth, stillbirth, placental abruption and small for gestational age were 0.622 to 0.683. Flagging the top 10% by gestational diabetes risk found over 40% of cases; lifestyle counselling was modelled as cost-effective. Kaiser models applied to 6105 All of Us pregnancies did not consistently match locally trained models, even after partial retraining.

Why it is interesting: Most prior models used specialised data or unrepresentative cohorts, whereas this audit uses only routinely collected records. It shows which complications such data can predict, and that models need local tuning in a new health system.

Source
Communications Health, 13 November 2026
Paper
Evaluation of machine learning-based prediction of pregnancy complications and adverse birth outcomes using routine data
Authors
Parent C, Moraitaki C, Ngo A, Greenberg MB, Zhu Y, Ghassemi M
DOI
10.1038/s44528-026-00051-0
Full text
Open copy (open-access copy, PubMed Central)
Type
Peer-reviewed article
Design
Retrospective population-based cohort, 442,760 pregnancies (Kaiser Permanente Northern California, 2011-2021), XGBoost models for seven outcomes at six-week intervals, external transfer cohort n=6,105
Verdict
Refines prior evidence
Driver
AI in medicine

Other drivers in Data, models and systems