Science

Science, technology and the emergence of AI have converged to make precision medicine for acute care a reality. This is how we deliver healthcare tailored to biology.

Research & Development

Evidence

Clinical Trials

The Sepsis ImmunoScore® Predicts Sepsis, Mortality, and Deterioration Better than Clinical Scores and Widely Available Biomarkers

Read More →

Products

Prenosis is developing a pipeline of integrated diagnostics and therapeutics grounded in each patient's unique biological profile, using deep data science and AI to give clinicians tools that reflect the true complexity of critical disease.

Immunix®

Sepsis ImmunoScore®

BioAegis Therapeutics Announces Collaboration with Prenosis to Advance AI-Driven Precision Medicine in Inflammatory Disease

Read More →

Resources

Blogs, articles, videos, and podcasts exploring how precision medicine and AI are changing the way clinicians detect and treat critical conditions.

Resource Library

Events

Treating the Patient, Not the Disease

Read More →

Newsroom

Read the latest Prenosis peer-reviewed articles and items from scholarly journals.

Newsroom

Sepsis ImmunoScore Outperforms Conventional Sepsis Tools

Read More →

About

As a biology-based technology company, we combine data science with AI to transform how critical conditions are diagnosed and treated.

Mission, Vision & Values

Careers

Leadership

Awards

LifeBridge Health deploys Prenosis’ FDA-authorized AI tool to speed sepsis detection

Read More →

Sepsis ImmunoScore®

The first FDA-authorized AI diagnostic to support early sepsis risk assessment

Sepsis ImmunoScore leverages biomarkers, clinical data, and AI trained on thousands of patients to holistically evaluate an individual patient’s chances of having or developing sepsis within 24 hours.

With direct integration to hospital EMRs, the report provides full transparency into how the score was calculated, right in the clinical workflow.

Sepsis ImmunoScore results in four risk categories that are associated with:

Chance of Sepsis Within 24 Hours

In-Hospital Mortality

Hospital Length of Stay

ICU Transfer Within 24 Hours

Vasopressors Used Within 24 Hours

Mechanical Ventilation Within 24 Hours

Solving for the AI Black Box

Most AI solutions for sepsis hide the algorithms that inform their predictions.

With a graphical interface that displays 22 decision parameters, Sepsis ImmunoScore builds clinician trust in AI to assist in decision-making. Read the evidence supporting how and why our tool works.

The Sepsis ImmunoScore® Predicts Sepsis, Mortality, and Deterioration Better than Clinical Scores and Widely Available Biomarkers

Read More →

Interpretability of an FDA-authorized AI/ML sepsis diagnostic tool improved by SHAP values

Read More →

Challenges Facing Developers of Diagnostic Tests for Sepsis: A Report From Sepsis Alliance and the Infection Management and Sepsis Collaborative Community

Read More →

Frequently Asked Questions

We answered the most common questions clinicians and hospital leaders ask about Sepsis ImmunoScore.

What is Sepsis ImmunoScore?

Sepsis ImmunoScore is an FDA-Authorized Artificial Intelligence/Machine Learning (AI/ML)-Based Software that identifies patients at risk for having or developing sepsis.

It uses up to 22 inputs from the electronic health record (EHR) to:

  • Generate a quantitative risk score
  • Assign patients to one of four discrete risk stratification categories: low, medium, high, or very high

These categories reflect increasing risk of sepsis and support clinical decision-making.

How is Sepsis ImmunoScore used in clinical practice?

Sepsis ImmunoScore is used alongside lab results and clinical evaluation to help assess a patient’s risk of sepsis or progression within 24 hours.

It is intended to be used:

  • To support assessment of sepsis risk in patients admitted to the Emergency Department or hospital with suspected sepsis, as indicated by the order of a blood culture
  • In conjunction with other laboratory findings, not as a replacement of clinical judgment
  • By licensed healthcare providers
Who is Sepsis ImmunoScore intended for?

Sepsis ImmunoScore is intended for use in:

  • Adult patients (18 years or older) admitted to the Emergency Department or hospital with suspected sepsis as indicated by the order of a blood culture

Sepsis ImmunoScore has not been validated for use in certain inpatient settings such as ICU or Labor and Delivery units.

What is an AI biomarker?

The FDA, through its BEST (Biomarkers, EndpointS, and other Tools) Glossary developed with the NIH, defines a biomarker as:

“A defined characteristic that is measured as an indicator of normal biological processes, pathogenic processes, or responses to an exposure or intervention (including therapeutic interventions).”

The FDA recognizes multiple categories of biomarkers, including Diagnostic, Prognostic, Predictive, Pharmacodynamic/Response, Monitoring, Safety, and Susceptibility/Risk.

Prenosis recognized the need for a new category: the AI-Biomarker.

An AI-Biomarker is:

“A defined, algorithm-derived characteristic generated through artificial intelligence or machine learning methods that integrate and analyze one or more data modalities (e.g., molecular, physiologic, imaging, electronic health record, or clinical data) to provide a reproducible measure indicative of normal biological processes, pathogenic processes, or responses to an exposure or intervention.”

Put more simply:

An AI biomarker is a computed biological signal that combines patterns across multiple types of patient data, such as:

  • Clinical data
  • Laboratory values
  • Physiological measurements
  • Other biomarkers

The result is a reproducible score that serves the same functions as a traditional biomarker—detecting disease, predicting outcomes, or assessing treatment response.

Where is Sepsis ImmunoScore available?

Sepsis ImmunoScore is currently approved and available solely in the United States. Additional markets are under exploration.

How does Sepsis ImmunoScore work?

Sepsis ImmunoScore inputs up to 22 parameters from the electronic health record (EHR), calculates a sepsis risk score using a locked algorithm, and outputs the Sepsis Patient View, a screen that displays the risk score along with the contributing clinical parameters and their impact on the score.

The software integrates directly within the EHR for seamless workflow integration.

View our interactive demonstration to see how it works.

What results does Sepsis ImmunoScore provide?

Sepsis ImmunoScore calculates a patient’s risk of sepsis and assigns it to one of four risk categories: low, medium, high, or very high risk.

Each category indicates a patient’s risk of having sepsis or developing sepsis within the next 24 hours.

These risk categories are highly correlated with clinical deterioration within 24 hours, including ICU admission, vasopressor use, mechanical ventilation, in-hospital mortality, and length of stay.

As a multi-band test, Sepsis ImmunoScore provides more actionable insight than a single positive/negative result.

 

What data does Sepsis ImmunoScore use to generate the sepsis risk score?

A: Sepsis ImmunoScore analyzes up to 22 patient parameters—including demographics, vital signs, lab values, and biomarkers—pulled directly from the EHR system:

  • Age
  • Systolic Blood Pressure
  • Diastolic Blood Pressure
  • Temperature
  • Respiratory Rate
  • Heart Rate
  • Blood Oxygen Saturation
  • White Blood Cell Count
  • Lymphocyte Count
  • Neutrophil Count
  • Platelet Count
  • Blood Urea Nitrogen
  • Creatinine
  • Potassium
  • Chloride
  • Total Carbon Dioxide
  • Sodium
  • Albumin
  • Bilirubin
  • Procalcitonin
  • C-Reactive Protein
  • Lactate
How is Sepsis ImmunoScore different from other sepsis technologies?

Early warning systems function primarily as alert tools, while Sepsis ImmunoScore is an FDA-authorized diagnostic device. These approaches are fundamentally different.

Early warning systems passively monitor for signs of deterioration, whereas Sepsis ImmunoScore actively engages providers when sepsis is suspected to support diagnostic and treatment decisions.

Sepsis ImmunoScore combines up to 22 parameters from the EHR using a validated algorithm to deliver a quantitative, individualized risk score. When used alongside early warning systems, Sepsis ImmunoScore improves the identification of true sepsis patients and could reduce false alerts.

Because it does not rely on physical test kits or analyzers, is it not impacted by supply chain disruptions. And because it integrates seamlessly with the EHR, it can scale across an entire hospital or health system very quickly. Providers simply order it and view results like any other lab test, all within their existing clinical workflow.

For more information:
[link to Epic paper when done]
[link to comparators paper if/when published]

Is Sepsis ImmunoScore externally validated?

Yes, Sepsis ImmunoScore is a Software as Medical Device (SaMD), and the first and only FDA-authorized AI diagnostic tool for sepsis through the De Novo pathway.

What evidence supports Sepsis ImmunoScore?

Sepsis ImmunoScore is supported by clinical evidence, including FDA-reviewed data and peer-reviewed studies.

Two published studies include:

Bhargava A. et al. Validation of a Machine Learning Diagnostic Tool for the Prediction of Sepsis and Critical Illness. https://www.medrxiv.org/content/10.1101/2024.05.06.24306954v1.full-text

Bhargava A. et al. FDA-Authorized AI/ML Tool for Sepsis Prediction: Development and Validation. https://ai.nejm.org/doi/full/10.1056/AIoa2400867

How accurate is Sepsis ImmunoScore in predicting sepsis?

A 2024 study published in the New England Journal of Medicine AI (NEJM AI) demonstrated strong diagnostic performance for identifying sepsis and predicting clinical deterioration.

Does Sepsis ImmunoScore continuously learn?

The deployed Sepsis ImmunoScore algorithm is a locked model and does not perform continuous learning. This ensures consistent, validated performance across all clinical environments.

How is algorithm drift managed for Sepsis ImmunoScore?

Prenosis maintains a robust post-implementation monitoring program of Sepsis ImmunoScore to evaluate model performance and detect potential algorithm drift. Ongoing review ensures continued accuracy and alignment with regulatory and clinical performance standards.

What data was used to train Sepsis ImmunoScore?

Data used to train and validate the AI/ML model comes from Prenosis’ proprietary dataset and biobank, comprised of de-identified data from multiple hospital sites, including:

  • Prospectively collected clinical data (EHR data)
  • Time-series biological samples
  • Biomarker measurements

The dataset includes adult patients admitted to the Emergency Department or hospital with suspected sepsis, as defined by the first order of a blood culture.

What are the key clinical benefits of Sepsis ImmunoScore?

Sepsis ImmunoScore delivers a clear risk score based on each patient’s individual biology to support clinical decision-making.

  • Quantitative, biology-based risk stratification
    • Sepsis ImmunoScore combines up to 22 patient parameters to deliver a clear, quantitative score that helps clinicians quickly assess sepsis risk.
  • Early prediction of adverse outcomes
    • The tool predicts sepsis onset and related outcomes, including ICU admission, mechanical ventilation, vasopressor use, in-hospital mortality, and length of stay.
  • Support for guideline-aligned sepsis care
    • By identifying at-risk patients sooner, Sepsis ImmunoScore can help hospitals improve SEP-1 bundle performance, standardize sepsis management, and deliver faster, evidence-based intervention.
Is Sepsis ImmunoScore reimbursed by CMS or commercial payers?

Sepsis ImmunoScore is currently included within the Diagnosis-Related Group (DRG) for sepsis, with no separate reimbursement.

However, as of October 1, 2025, a new ICD-10-PCS code became available for AI sepsis diagnostic tools, including Sepsis ImmunoScore.

For more information, review our Hospital Billing and Coding Guide.

1. Rudd, K. E. et al. Global, regional, and national sepsis incidence and mortality, 1990–2017: analysis for the Global Burden of Disease Study. The Lancet 395, 200–211 (2020).

2. Human factors study performed for FDA De Novo Approval.