Study: Machine learning tool predicts preeclampsia risk

A retrospective, multisite cohort study found a machine-learning tool could predict preeclampsia in late-stage pregnancy using routinely available clinical and laboratory data. The study, published in JAMA Network Open, suggests the model offers opportunities for earlier intervention and is adaptable across diverse health care settings. Most tools for diagnosing preeclampsia currently focus on risk prediction in early pregnancy or rely on biomarkers or imaging tests. The study included 58,839 pregnancies from October 2020 to May 2025 at three New York-Presbyterian hospitals, with prediction performance peaking at 34 weeks. The patients who developed preeclampsia were older and more often Black. Preeclampsia is a major cause of maternal and infant morbidity and mortality, affecting 2% to 8% of pregnancies worldwide. (MedPage Today article, 3/6/26)