Comparison of Sepsis Prediction Algorithms

Prospective Evaluation of Sepsis Prediction Algorithms in a Multi-Hospital Healthcare System

Sepsis is a severe response to infection resulting in organ dysfunction and often leading to death. More than 1.5 million people get sepsis every year in the U.S., and 270,000 Americans die from sepsis annually. Delays in the diagnosis of sepsis lead to increased mortality. Several clinical decision support algorithms exist for the early identification of sepsis. The research team will compare the performance of three sepsis prediction algorithms to identify the algorithm that is most accurate and clinically actionable. The algorithms will run in the background of the electronic health record (EHR) and the predictions will not be revealed to patients or clinical staff. In this current evaluation study, the algorithms will not affect any part of a patient's care. The algorithms will be deployed across the Emory healthcare system on data from all patients presenting to the emergency department.

Study Overview

• Status: Not yet recruiting
• Conditions: Sepsis
• Intervention / Treatment: Other: Epic Sepsis Model Version - 1; Other: Epic Sepsis Model Version - 2; Other: Emory Sepsis Model

Detailed Description

The primary goal of this study is to prospectively evaluate three sepsis prediction algorithms that are embedded in the EHR. The models will be deployed in a "shadow" mode, and the results will not be displayed to the treatment team during this study. Two of the algorithms are proprietary algorithms of the EHR provider (Epic). The third algorithm is an internally developed, open-source algorithm.

The algorithms will compute the probability of sepsis at periodic intervals and will continue to run on a patient's data until the patient's discharge, death, or upon initiation of intravenous antibiotics (at which point there is an indirect record of clinical suspicion of an infection).

• Study Type: Observational
• Enrollment (Estimated): 1200

Contacts and Locations

• Study Contact: Name: Sivasubramanium Bhavani, MD; Phone Number: 404-712-2970; Email: sivasubramanium.bhavani@emory.edu

Study Locations

• United States
  • Georgia
    • Atlanta, Georgia, United States, 30308
      • Emory Saint Joseph's Hospital
    • Atlanta, Georgia, United States, 30308
      • Emory Midtown Hospital
      • Contact: Sivasubramanium Bhavani, MD; Phone Number: (404) 501-1000; Email: sivasubramanium.bhavani@emory.edu
    • Atlanta, Georgia, United States, 30322
      • Emory Hospital
    • Atlanta, Georgia, United States, 30322
      • Emory Healthcare System
    • Decatur, Georgia, United States, 30033
      • Emory Decatur Hospital
      • Contact: Sivasubramanium Bhavani, MD; Phone Number: 404-712-2970; Email: sivasubramanium.bhavani@emory.edu
    • Johns Creek, Georgia, United States, 30097
      • Emory Johns Creek Hospital
    • Lithonia, Georgia, United States, 30058
      • Emory Hillandale Hospital
      • Contact: Sivasubramanium Bhavani, MD; Phone Number: (404) 501-8000; Email: sivasubramanium.bhavani@emory.edu

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: Adult; Older Adult
• Accepts Healthy Volunteers: No

• Sampling Method: Non-Probability Sample

Study Population

All adult patients admitted through the ED from the time they arrive at the hospital until discharge, death, or initiation of intravenous antibiotics.

Description

Inclusion Criteria:

• All adult patients admitted through the ED

Exclusion Criteria:

• None

Study Plan

How is the study designed?

Number of groups / cohorts

1

Cohorts and Interventions

Group / Cohort

Intervention / Treatment

ED Patients; All adult patients presenting to Emergency Departments (ED) in the Emory Healthcare system

Other: Epic Sepsis Model Version - 1; The Epic Sepsis Model (ESM) version 1, a proprietary sepsis prediction model.; Other Names: Proprietary Epic sepsis algorithm -1; Other: Epic Sepsis Model Version - 2; The Epic Sepsis Model (ESM) version 2, a proprietary sepsis prediction model.; Other Names: Proprietary Epic sepsis algorithm -2; Other: Emory Sepsis Model; Emory internal algorithm; Other Names: Emory Sepsis Algorithm

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Patient hospitalization-level area under curve (AUC) for identification of sepsis,	Definition of Sepsis using the Centers for Disease Control and Prevention (CDC) Adult Sepsis Surveillance.	Duration of hospital stay (until discharge or death), an expected average of 30 days

Secondary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Sensitivity, specificity, and Positive and Negative Predictive Value of algorithms	Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV).	Duration of hospital stay (until discharge or death), an expected average of 30 days
Lead time to antibiotic administration	The time between the initial deployment of the alert in patients confirmed to have sepsis (ture positives) and the physician's ordering of intravenous antibiotic therapy.	Duration of hospital stay (until discharge or death), an expected average of 30 days
Percent expected increase in unnecessary antibiotics	Percent of patients who were incorrectly identified as having sepsis (false positives), and received antibiotics.	Duration of hospital stay (until discharge or death), an expected average of 30 days
Number needed to screen	The number of alerts that would need to be processed to find one true positive sepsis.	Duration of hospital stay (or death), an expected average of 30 days
Time-horizon based AUCs	AUCs will be calculated at 3 pre-specified time horizons.	4 hours, 8 hours, and 24 hours
Accuracy and calibration by subgroup	The AUC and calibration curves will be compared by sex and race to ensure predictive accuracy is equal across subgroups.	Duration of hospital stay (until discharge or death), an expected average of 30 days
Number of Total and false alert burden	The number of Total and false alert burden cumulative across all study patients over the study period	Duration of hospital stay (until discharge or death), an expected average of 30 days

Collaborators and Investigators

• Sponsor: Emory University
• Investigators: Principal Investigator: Sivasubramanium Bhavani, MD, Emory University

Study record dates

Study Major Dates

• Study Start (Estimated): June 1, 2026
• Primary Completion (Estimated): December 1, 2026
• Study Completion (Estimated): December 1, 2026

Study Registration Dates

• First Submitted: July 5, 2023
• First Submitted That Met QC Criteria: July 5, 2023
• First Posted (Actual): July 13, 2023

Study Record Updates

• Last Update Posted (Actual): January 7, 2026
• Last Update Submitted That Met QC Criteria: January 5, 2026
• Last Verified: December 1, 2025

More Information

Terms related to this study

• Keywords: Infection; Emergency Department; Prediction; Algorithm
• Additional Relevant MeSH Terms: Pathologic Processes; Disease Attributes; Systemic Inflammatory Response Syndrome; Inflammation; Pathological Conditions, Signs and Symptoms; Emergencies; Sepsis; Infections
• Other Study ID Numbers: STUDY00005958; 2024P008316 (Other Identifier: Emory IRB)

Plan for Individual participant data (IPD)

• Plan to Share Individual Participant Data (IPD)?: YES
• IPD Plan Description: The investigators will share individual participant data that underlie the results reported in an article, after deidentification (text, tables, figures, and appendices).
• IPD Sharing Time Frame: Data will be shared beginning 6 months after publication, without a specified end date.
• IPD Sharing Access Criteria: To gain access, data requestors will need to sign a data access agreement.
• IPD Sharing Supporting Information Type: STUDY_PROTOCOL

Drug and device information, study documents

• Studies a U.S. FDA-regulated drug product: No
• Studies a U.S. FDA-regulated device product: No
• product manufactured in and exported from the U.S.: No