Related Experiment Video
Updated: Aug 11, 2026
07:31
Implementation of a Real-Time Psychosis Risk Detection and Alerting System Based on Electronic Health Records using CogStack
Published on: May 15, 2020
A Host-Clinical Integrated Prediction Model for Early Sepsis Risk in Diabetic Patients with Acute Infections:
Xilong Pan1, Zhiyuan Xu2, ChanJuan Zhuo1
1Department of Clinical Laboratory, Dongsheng Hospital, Zhongshan, Guangdong, People's Republic of China.
Journal of Inflammation Research
|August 10, 2026
Summary
This study developed a sepsis risk model for diabetic patients with infections. The model accurately predicts sepsis risk, outperforming existing scores and offering a novel tool for early detection.
Area of Science:
- Medical research
- Clinical informatics
- Infectious disease
Background:
- Diabetic patients have a higher risk of infection and sepsis progression.
- Early sepsis detection is crucial for improving patient outcomes.
Purpose of the Study:
- To develop and validate a sepsis risk prediction model for hospitalized diabetic patients with acute infections.
- To compare the model's performance against established scoring systems (NEWS2, SIRS, qSOFA).
Main Methods:
- Retrospective analysis of 604 diabetic patients with acute infections.
- Development of a predictive model using penalized regression and Firth logistic regression.
- Internal validation and comparison with NEWS2, SIRS, and qSOFA.
Main Results:
- The final model integrated temperature dysregulation, nausea/vomiting, fatigue, albumin, and procalcitonin.
- The model achieved a corrected AUC of 0.915 and outperformed NEWS2, SIRS, and qSOFA.
- Consistent performance was observed across subgroups, with superior clinical utility and cost-effectiveness.
Conclusions:
- The developed host-clinical integrated model serves as a novel tool for early sepsis risk warning in diabetic patients.
- This approach offers insights into sepsis risk prediction in specific populations.
- External validation is recommended prior to clinical implementation.