Innovative, Patient-Specific, Machine Learning Technology Can Improve Sepsis Care

Study in a Sentence: Researchers combined patient-specific samples with machine learning technology to develop a bedside system that can predict the likelihood that a patient with suspected sepsis will worsen and helps identify necessary interventions more quickly.

Healthy for Humans: Sepsis is a life-threatening condition that occurs when the immune system has a dysfunctional response to an infection. Early patient-specific interventions may significantly improve long-term outcomes and prevent death. Machine learning, lab-on-a-chip technologies, using blood samples from patients with suspected sepsis, can predict at the bedside whether a patient’s condition is likely to worsen over the next 24 hours—helping prioritize those who need urgent interventions.

Redefining Research: Existing treatment of sepsis relies on measuring the severity of clinical symptoms rather than addressing individualized patient needs, and current diagnostic methods often fail to identify patients at risk of developing sepsis at first clinical presentation. Additionally, many diagnostic tests are only available at well-equipped facilities, making them less accessible for vulnerable populations. In addition to patient-specific predictive capabilities, lab-on-a-chip technologies enable portable, point-of-care testing that may advance sepsis care across a wide range of health care settings.