The Sepsis ImmunoScore is the first sepsis artificial intelligence diagnostic tool authorized by the FDA and uses patient data to assess risk within 24 hours.
RT’s Three Key Takeaways:
- Early Sepsis Detection: Sinai Hospital has integrated Sepsis ImmunoScore, the first artificial intelligence diagnostic tool authorized by the FDA to assess sepsis risk up to 24 hours in advance.
- Clinical Workflow Integration: The software evaluates 22 clinical and laboratory measurements directly within the electronic medical record to guide decision-making without adding clinical steps.
- Substantial Sepsis Mortality: According to the CDC, sepsis affects at least 1.7 million adults in the US each year and contributes to one in three hospital deaths.
Clinical teams at Sinai Hospital (Baltimore) have begun using a new artificial intelligence diagnostic tool designed to flag patients at risk of developing sepsis, according to LifeBridge Health. The system, Sepsis ImmunoScore, is the nation’s first sepsis artificial intelligence diagnostic tool authorized by the USFDA.
Developed by Prenosis, the tool analyzes up to 22 clinical and laboratory parameters to detect patterns indicating whether an adult patient is developing or is at risk of developing sepsis within the next 24 hours, the news release stated. The software is embedded directly within Sinai Hospital’s electronic medical record, enabling clinicians to view results during routine workflow alongside standard laboratory results, physical evaluations, and clinical judgment.
Addressing Diagnostic Delays
Sepsis occurs when the body’s reaction to an infection damages its own tissues and organs, leading to rapid health declines, according to the CDC. Data from the agency indicate that at least 1.7 million adults in the US develop sepsis annually, resulting in at least 350,000 hospital deaths or discharges to hospice. Additionally, the CDC reports that one in three patients who dies in a hospital setting has sepsis.
Because early symptoms can mimic numerous other acute conditions and definitive blood cultures can take several days to finalize, establishing a prompt diagnosis is a common challenge, according to the health system.
“Sepsis is notoriously difficult to diagnose early because its signs can mimic many other serious illnesses,” said Dr Gregory Corcoran, chief of emergency medicine at LifeBridge Health, in a news release. “Anything that helps physicians identify patients at risk sooner can make a meaningful difference in how quickly treatment begins.”
Corcoran noted that early diagnostic insight provides value beyond the emergency department (ED) setting.
“The greatest benefit of this technology extends beyond the emergency department,” said Dr Gregory Corcoran, chief of emergency medicine at LifeBridge Health. “By providing additional diagnostic confidence early, it helps physicians make more informed decisions as patients transition from the ED to inpatient care. That can reduce unnecessary antibiotic treatment for patients who don’t have sepsis while helping us identify those who do before their condition worsens.”
From Research to Clinical Application
Sinai Hospital participated in the initial development of the platform prior to its commercial deployment, according to the release. The hospital served as one of 10 research centers across the US that supplied patient data to a biobank of more than 100,000 biological samples used to train and validate the diagnostic algorithm. The initiative was coordinated across the Department of Research at LifeBridge Health, the Sinai Center for Thrombosis Research, and multiple clinical units.
“Our physicians and researchers helped lay the foundation for this new sepsis technology, and it’s gratifying that now our clinical teams are using it to help identify one of medicine’s most time-sensitive conditions,” said Dr Charles Albrecht III, chief medical officer at Sinai Hospital. “Earlier recognition of sepsis can make a critical difference for patients.”
LifeBridge Health stated that it plans to assess the impact of the software on clinical endpoints, including treatment delivery, length of stay, patient readmissions, and mortality, before determining whether to expand the tool to additional healthcare facilities across its network.
