The US Food and Drug Administration (FDA) has granted marketing authorisation for Prenosis' AI tool Sepsis ImmunoScore, to diagnose sepsis.

This authorisation, achieved through the De Novo pathway, represents the first-ever FDA approval of an AI/machine learning (ML) software as a medical device (SaMD) specifically for sepsis diagnosis and prediction.

Sepsis, a condition triggered by an overactive response to infection, poses a substantial burden on the US healthcare system, with billions of dollars in costs and millions of deaths annually.

The Sepsis ImmunoScore utilises AI to combine biomarkers and clinical data, providing a comprehensive risk assessment for the onset or progression of sepsis within 24 hours of patient evaluation in emergency or hospital settings.

It evaluates 22 varied parameters to calculate a risk score and categorise patients into four discrete risk levels.

These categories are indicative of potential patient outcomes, including hospital stay duration, mortality, and the need for critical care interventions.

Integrated directly into a hospital’s electronic medical records (EMRs), the dual diagnostic and predictive software presents a transparent view of how each parameter contributes to the patient's sepsis score.

This transparency is crucial for clinician trust and fosters a synergistic relationship between healthcare providers and the AI system, leading to quicker and more effective treatment decisions.

Prenosis co-founder and CEO Bobby Reddy Jr said: “FDA authorisation of a sepsis diagnostic tool with significant predictive power is a landmark event for people that could ever be at risk of sepsis at some point in their lives.

“Until now, there was no other FDA-authorised AI diagnostic for sepsis, which is why the Sepsis ImmunoScore had to be granted marketing authorisation through the De Novo pathway. FDA authorisation offers yet another important piece of evidence of the potential of the Sepsis ImmunoScore to improve care.”