Related Experiment Video
Updated: May 28, 2025
Implementation of a Real-Time Psychosis Risk Detection and Alerting System Based on Electronic Health Records using CogStack
Published on: May 15, 2020
Development and Validation of a Machine Learning Model for Early Prediction of Sepsis Onset in Hospital Inpatients
Pierre-Elliott Thiboud1, Quentin François1, Cécile Faure1
1PREVIA MEDICAL, 69007 Lyon, France.
Insights
A new machine learning algorithm, the SEPSI Score, accurately predicts early sepsis onset in hospitalized patients. This tool outperforms existing systems, detecting sepsis hours before clinical confirmation, aiding timely patient management.
Area of Science:
- Medical Informatics
- Machine Learning in Healthcare
- Critical Care Medicine
Background:
- Sepsis causes 11 million deaths globally, necessitating early detection tools.
- Hospitalized patients are at risk, making early prediction a public health priority.
- Current sepsis prediction methods require improvement for timely intervention.
Purpose of the Study:
- To develop and validate a machine learning algorithm for early sepsis prediction.
- To assess the performance of the SEPSI Score against existing sepsis scoring systems.
- To evaluate the potential of the SEPSI Score in all hospital departments.
Main Methods:
- Retrospective collection of sepsis predictors from 45,127 patients.
- Development of the SEPSI Score using a gradient boosted trees approach.
- Validation on 5270 patient stays, including 121 sepsis cases.
Main Results:
- SEPSI Score achieved a positive predictive value of 0.610, significantly higher than SOFA (0.174).
- Area under the precision-recall curve was 0.738 for SEPSI Score, outperforming SOFA (0.174).
- High sensitivity (0.845) and specificity (0.987) were reported; sepsis detected up to 48 hours earlier.
Conclusions:
- The SEPSI Score accurately predicts early sepsis onset, surpassing current scoring systems.
- This predictive tool can be implemented across hospital departments for early sepsis management.
- Further assessment is needed to evaluate the impact of SEPSI Score on morbidity and mortality.
Abstract:
Background: With 11 million sepsis-related deaths worldwide, the development of tools for early prediction of sepsis onset in hospitalized patients is a global health priority. We developed a machine learning algorithm, capable of detecting the early onset of sepsis in all hospital departments. Methods: Predictors of sepsis from 45,127 patients from all departments of Valenciennes Hospital (France) were retrospectively collected for training. The binary classifier SEPSI Score for sepsis prediction was constructed using a gradient boosted trees approach, and assessed on the study dataset of 5270 patient stays, including 121 sepsis cases (2.3%). Finally, the performance of the model and its ability to detect early sepsis onset were evaluated and compared with existing sepsis scoring systems. Results: The mean positive predictive value of the SEPSI Score was 0.610 compared to 0.174 for the SOFA (Sepsis-related Organ Failure Assessment) score. The mean area under the precision-recall curve was 0.738 for SEPSI Score versus 0.174 for the most efficient score (SOFA). High sensitivity (0.845) and specificity (0.987) were also reported for SEPSI Score. The model was more accurate than all tested scores, up to 3 h before sepsis onset. Half of sepsis cases were detected by the model at least 48 h before their medically confirmed onset. Conclusions: The SEPSI Score model accurately predicted the early onset of sepsis, with performance exceeding existing scoring systems. It could be a valuable predictive tool in all hospital departments, allowing early management of sepsis patients. Its impact on associated morbidity-mortality needs to be further assessed.
