Researchers used electronic health records (EHRs) and advanced machine learning methodologies to predict neonatal outcomes from maternal conditions. The study used data from 32,354 mother-newborn dyads to train and validate the model and found it outperformed currently used EHR-based clinical risk scores.
The study identified previously unreported associations between maternal conditions, such as anemia, certain medication exposures, and social determinants of health, with neonatal outcomes. A longitudinal clinical risk calculator to predict neonatal outcomes was developed in the study , enabling individualized care and aiding in the identification of suitable patients for clinical trials. It can be accessed at https://maternal-child-health-associations.shinyapps.io/shiny_app/
Ref: De Francesco D, Reiss JD, Roger J, et al. Data-driven longitudinal characterization of neonatal health and morbidity. Science Translational Medicine. 2023;15(683). doi:https://doi.org/10.1126/scitranslmed.adc9854