In maternal and newborn health care, a pressing challenge looms large – neonatal sepsis. This dangerous infection claims the lives of countless newborns each year, devastating families and other loved ones.

The World Health Organization (WHO) identifies neonatal infections as a leading cause of infant mortality, responsible for nearly a quarter of all newborn deaths worldwide. Neonatal sepsis arises when the newborn’s immune system response to infection leads to widespread inflammation and organ dysfunction. It’s particularly prevalent in areas with low resources and limited access to quality health care.

Sepsis can be hard to diagnose and is a global health issue that demands attention and innovative solutions.

Human expertise and AI join forces

Could the use of AI to monitor vitals and predict the likelihood of sepsis save the lives of premature newborns?  A team of doctors from Portugal set out to come up with the answer to that question.

Team IN-STEP, titled and stylized as “savINg liveS fighTing nEonatal sePsis” (IN-STEP), developed a solution that augments human expertise and AI-generated insights into medical interventions and disease diagnosis.

Team IN-STEP used SAS® Viya® to develop AI-powered models for early detection, timely intervention and education to help communities and providers understand sepsis more. Health care providers are great at analyzing data and patterns for diagnosis, but human error is inevitable. Also, relying on results from laboratory tests alone isn’t enough.

“When we have a newborn with suspected sepsis, we value mostly the clinical assessment. Vital signs such as heart rate, respiratory rate, blood pressure and peripheral perfusion are important to us,” said Dr. João Virtuoso, pediatrician and team lead on the project. “We also link to the laboratory parameters, namely the leukogram. Hemoculture is also data we use, but that always takes some more time to be positive. Integrating everything into a single variable would be very useful and help us identify sepsis earlier and start antibiotics as soon as possible."

Team IN-STEP’s solution fills that gap by creating models to wrangle those many variables in disease diagnosis.

“Because newborns are often asymptomatic and the assessment is often nonspecific or common to other pathologies, we often do not know what disease that newborn may have,” said Dr. Madalena Lopo Tuna, pediatrician and a team member on the project. “On the other hand, results from laboratory tests alone will not help, hence the great advantage of having a model that integrates all of the variables.”

Team IN-STEP was one of the winners at the 2023 SAS Hackathon for their efforts in developing a solution for the early detection of neonatal sepsis.

Bring your ideas and join us

Interested in other innovations from the SAS Hackathon?

Come for a visit to see what teams have been able to create – from concepts to fully working solutions. And think about how you can create the next big thing in your industry by joining us in 2024.


About Author

Caslee Sims

I'm Caslee Sims, writer and editor for SAS Blogs. I gravitate toward spaces of creativity, collaboration and community. Whether it be in front of the camera, producing stories, writing them, sharing or retweeting them, I enjoy the art of storytelling. I share interests in sports, tech, music, pop culture among others.

Leave A Reply

Back to Top