Researchers at Chicago-based Northwestern University and Ann & Robert H. Lurie Children’s Hospital of Chicago developed AI models that identify children with high risk of developing sepsis within 48 hours using electronic health record data.
The models were trained on data from the first four hours of emergency department care, before organ dysfunction developed, according to the study published Oct. 13 in JAMA Pediatrics. Researchers validated the models using retrospective data from five health systems contributing to the Pediatric Emergency Care Applied Research Network. The training data spanned January 2016 to February 2020, with validation from 2021 to 2022.
This study is the first to apply the Phoenix Sepsis Criteria to pediatric AI modeling. It excluded children who already had sepsis at arrival or during the first hours of care, focusing instead on early identification for timely intervention.
Predictive features included triage score, heart and respiratory rates and preexisting conditions such as cancer. Researchers emphasized that future clinical use should combine the models with clinician judgment to avoid overidentifying low-risk patients.
At the Becker's 11th Annual IT + Revenue Cycle Conference: The Future of AI & Digital Health, taking place September 14–17 in Chicago, healthcare executives and digital leaders from across the country will come together to explore how AI, interoperability, cybersecurity, and revenue cycle innovation are transforming care delivery, strengthening financial performance, and driving the next era of digital health. Apply for complimentary registration now.