ML Model May Predict Preeclampsia Risk

Dynamic EHR-based approach may enable earlier intervention in late pregnancy.

Share

1

A machine learning model may predict short-term preeclampsia risk using electronic health record data from late pregnancy.
2

The study involved 58,839 pregnancies and demonstrated good discrimination in predicting preeclampsia onset.
3

Blood pressure was the most influential predictor, with laboratory values being more significant earlier in the third trimester.
4

The model's predictive accuracy improved as delivery approached, offering opportunities for earlier intervention.
5

Findings require prospective validation before clinical implementation due to the study's retrospective design.