Related Experiment Video
Updated: Jul 18, 2025
07:42
A Data-Driven Approach to Quantifying Immune States in Sepsis
Published on: February 7, 2025
249
Sepsis Prediction Model for Determining Sepsis vs SIRS, qSOFA, and SOFA.
Adam R Schertz1,2, Kristin M Lenoir3, Alain G Bertoni1,3
1Department of Internal Medicine, Atrium Health Wake Forest Baptist, Winston-Salem, North Carolina.
JAMA Network Open
|August 25, 2023
Summary
The Sepsis Prediction Model (SPM) shows higher accuracy than other tools but is too slow for timely sepsis screening. Further validation is needed for clinical use.
Area of Science:
- Healthcare Informatics
- Clinical Decision Support
- Sepsis Management
Background:
- The Sepsis Prediction Model (SPM) by Epic Systems generates a predicting sepsis score (PSS).
- Existing sepsis prediction tools include Systemic Inflammatory Response Syndrome (SIRS), Sequential Organ Failure Assessment (SOFA), and quick Sepsis-Related Organ Failure Assessment (qSOFA).
- The SPM has not been validated against these established tools.
Purpose of the Study:
- To evaluate the accuracy and timeliness of the SPM.
- To compare the SPM's performance against SIRS, qSOFA, and SOFA.
Main Methods:
- Retrospective cohort study of adult hospital admissions in a US health system (June 2019 - December 2020).
- Sepsis defined by antimicrobial use, blood cultures, and organ dysfunction (eSOFA).
- Analysis conducted from March 2021 to February 2023.
Main Results:
- Of 60,507 admissions, 1663 met sepsis criteria.
- SPM with PSS ≥ 8 showed the highest balanced accuracy (0.79).
- SPM's median time to positivity was 68 minutes, significantly slower than SIRS (7 minutes) and SOFA (28 minutes).
Conclusions:
- The SPM demonstrates superior balanced accuracy at higher thresholds compared to other models.
- However, the SPM's clinical utility as a sepsis screening tool is limited by its poor timeliness.
- Further research is needed to optimize SPM for rapid clinical application.
