The poster, “Evaluating vendor-derived pediatric sepsis predictive model in acute care settings,” was part of the AI Evaluation Showcase at the summit, held March 21-24 in Chicago, Ill.
The research investigates predictive models for identifying high-risk pediatric sepsis patients derived from electronic health record (EHR) data. Specifically, Dr. Yu and Dr. Liu examined an existing vendor-derived predictive model as the first phase of the research. They propose to correct the inefficiencies of this model in subsequent phases by developing a predictive model using Arkansas Children’s Hospital’s own EHR data.
“There is so much potential in applying artificial intelligence and machine learning to all the data available through EHR,” Dr. Yu said. “Done correctly, we can develop early alert systems that can assist our clinical care teams and save lives.”
This part of the research was an important first step, Dr. Liu said.
“Projects like this are especially vital in the pediatric setting, where informatics research is lagging behind the adult clinical environment,” Dr. Liu said. “It is past time that we bring the advances of informatics to pediatrics.”
AMIA – the American Medical Informatics Association – is a professional organization that connects a broad community of professionals and students interested in informatics.