LifeBridge Health’s Sinai Hospital in Baltimore uses artificial intelligence-powered sepsis diagnosis in clinical practice, bringing research-backed tools to patients’ bedsides.
Prenosis, a Chicago-based biology-based technology company, has developed Sepsis ImmunoScore, an AI-powered tool that can spot early signs of sepsis. The tool analyzes 22 parameters to identify patterns of a patient’s biological immune response, allowing clinicians to detect sepsis or progression to sepsis within 24 hours.
Sepsis ImmunoScore received initial approval from the U.S. Food and Drug Administration for software to diagnose sepsis in 2024. Prenosis’ solution was approved through the FDA’s De Novo pathway.
A study published in the New England Journal of Medicine AI found that the Sepsis Immune Score was highly accurate in diagnosing sepsis, while also predicting critical outcomes such as in-hospital mortality, length of stay to discharge, ICU admission, need for mechanical ventilation, and use of vasopressors.
Sinai Hospital in Baltimore is one of 10 research centers nationwide whose patient samples contributed to the 100,000-sample dataset that trains and validates the Prenosis algorithm. LifeBridge Health, which operates five hospitals and specialty care facilities in Maryland, has a long-standing relationship with the company as the health system’s researchers contributed to the basic research that made FDA approval of the tool possible.
“Our collaboration with Plenosis began with research, and we are pleased to see the research evolve into something that clinical teams can use directly. Adding new tools to the fight against sepsis truly reflects the type of forward-thinking innovation that defines our approach to patient care,” Charles Albrecht III, MD, chief medical officer at Sinai Hospital, said in a statement.
Sepsis is one of the costliest and most difficult diseases to manage in U.S. health care. At least 1.7 million adults and more than 18,000 children develop sepsis in the United States each year, according to data cited by the Centers for Disease Control and Prevention. At least 350,000 adults and more than 1,800 children who develop sepsis have died while hospitalized or been discharged to hospice care.
Sepsis is notoriously difficult to diagnose, and delays in recognition and treatment significantly worsen outcomes.
“Sepsis is a very difficult disease to diagnose. From the moment you arrive in the emergency room, there’s a lot of diagnostic uncertainty, and even after a patient is admitted to the hospital, it’s still uncertain whether this patient actually has an infection in their blood,” said Gregory Corcoran, M.D., director of emergency medicine at LifeBridge Health, in an interview with Fierce Healthcare. “The only way to know for sure is through blood culture results, which are still not 100% certain and often take several days to get results.”
This uncertainty could lead to overtreatment, where some patients receive unnecessary intravenous antibiotics, leading to antibiotic resistance, or undertreatment, where subtle cases of sepsis are missed, resulting in delayed treatment and worsening of patients’ symptoms, Corcoran said.
Sepsis consumes a huge portion of healthcare budgets, as patients require complex, multidisciplinary ICU care and long stays.
According to Dr. Bobby Reddy, Jr., CEO and co-founder of Prenosis, AI is well suited to address the complexity of sepsis diagnosis. “Sepsis is so complex and there are so many different biological subtypes of sepsis that it is important to collect enough information and the right information and use AI to compile that information to determine the appropriate treatment for the patient.”
Prenosis’ Sepsis ImmunoScore, built on the company’s Immunix precision medicine platform, integrates biomarkers and clinical data to present critical information in real time, helping clinicians act faster, potentially reducing damage to vital organs and improving patient survival rates, the company said. Integrates directly into the hospital’s electronic medical record system.
The company took an approach backed by research that shows biologically-based AI has the potential to change the way sepsis is diagnosed.
“We took the approach that sepsis is not just a disease,” Reddy told Fierce Healthcare. “We need to treat sepsis not as a single disease, but as a heterogeneous syndrome comprised of different types of biology that lead to sepsis. To better understand this highly heterogeneous and complex syndrome, Plenosis has Over the years, we’ve built up a repository of biological samples, blood samples taken from patients, and all the data we collect from patients – vitals, labs, medications, and other data – we analyze comorbidities, demographic information, and physical blood samples we take from patients, and we take those blood samples to our lab in Chicago, where we measure many deeper biological parameters that can’t be measured in the hospital.”
He added, “We’re literally talking about hundreds of parameters for each patient and then using AI to better analyze the different types of biology that make up sepsis. This is really precision medicine, a personalized approach.”
Reddy said it typically takes less than an hour from ordering a diagnostic test to getting results. “The Sepsis Score is very predictive because it gives clinicians an idea of the risk of developing sepsis within the next 24 hours,” he said.
Although antibiotics and other antimicrobial treatments can target the underlying infection, there are currently no treatments that directly treat the immune response that causes many sepsis cases.
Plenosis is also developing personalized immunotherapies that tailor treatment based on a patient’s specific immune profile, Reddy said.
Sinai Hospital in Baltimore brought the AI tool into clinical practice six weeks ago, and so far it has been used to test and diagnose 75 patients, Corcoran said.
“The reality of this tool is that the information you get by ordering this tool in the emergency department typically won’t change care in the ER. However, this tool increases the patient’s chances of success later on because it improves the diagnostic accuracy of the hospital physician who takes over the patient’s care,” Corcoran said. “The key to our success in the emergency department is that there are no workflow bottlenecks. If it’s easy to use for our team, doesn’t create extra steps, doesn’t create extra work for the ER provider, and we’re getting the benefit of diagnostic accuracy downstream, we know it’s good for the patient and we tend to use it.”
As the Sepsis Immune Score is used in clinical care, LifeBridge Health plans to evaluate several metrics, including ongoing evaluation of sepsis patient treatment and care, executives said.
Related: Prenosis signs new agreement for AI-powered sepsis risk tool
Plenosis is evaluating how the tool impacts hospitals’ compliance with the SEP-1 bundle (Medicare and Medicaid Services Severe Sepsis and Septic Shock Management Bundle), as well as how the tool improves length of stay, in-hospital mortality and 30-day readmission rates, Reddy said.
Building the infrastructure within Sinai Hospital will lay the foundation for implementing AI solutions across additional LifeBridge Health hospitals, said Martin Geshev, executive director of research at LifeBridge Health.
The company is also pursuing a broader strategy to build a precision medicine platform for lifesaving medicine. Prenosis’ biobank currently spans clinical data from 11 hospitals, 138,000 longitudinal specimens, and more than 35,000 patients.
The company recently signed an agreement with the Biomedical Advanced Research and Development Authority (BARDA) and completed a Series A funding round to expand the Immunix platform for diseases such as pneumonia, acute heart failure, acute respiratory distress syndrome, and acute kidney injury. Prenosis recently announced a strategic partnership with BioAegis Therapeutics to identify patient populations most likely to benefit from targeted therapies for inflammatory diseases.
“We see an opportunity to develop precision medicines for a wide range of indications across acute care settings, including sepsis, acute heart failure, pneumonia, acute kidney injury, and acute respiratory distress syndrome,” Reddy said. “What we’ve accomplished through building biobanks and collecting biological and clinical data is that we’ve discovered different features of the biology that span many of these different conditions.”
Reddy asserted that AI will revolutionize healthcare by promoting biologically-based precision medicine in acute care. “We are excited about the potential of AI to revolutionize this particular problem of understanding the complex biology of a patient and being able to deliver the right treatment to the right patient at the right time. In a hospital setting, timing is critical as patient biology can change rapidly in hours or days,” he said. “I think the realization of true precision medicine enabled by AI will lead to a fundamental transformation over the next five to 10 years.”
