Related Experiment Video
Updated: Jul 9, 2026
A Data-Driven Approach to Quantifying Immune States in Sepsis
Published on: February 7, 2025
Development and Validation of a Novel Predictive Model and Simplified Clinical Indicator for Mortality in Severe
1Department of Clinical Laboratory, The First Affliated Hospital of Anhui Medical University, Hefei, Anhui, People's Republic of China.
Purpose:
Severe fever with thrombocytopenia syndrome (SFTS) is a highly fatal infectious disease endemic in rural areas, underscoring the need for early prognostic risk stratification.
Patients And Methods:
This study aimed to develop and validate a prognostic model using readily available clinical indicators. We conducted a retrospective analysis of 260 SFTS patients (mortality rate 19.6%), integrating epidemiological characteristics, clinical symptoms and laboratory variables through a hybrid multi-model feature screening approach incorporating univariate Cox, LASSO, Random Forest, and XGBoost algorithms.
Results:
Four core risk factors for mortality were identified: age ≥ 56 years, platelet count ≤ 44 × 109/L, estimated glomerular filtration rate (eGFR) ≤ 53 mL/min/1.73m2, and activated partial thromboplastin time (APTT) ≥ 66.8 seconds. Based on these factors, we developed a comprehensive prognostic model and a simplified index termed AER (APTT-to-eGFR ratio). Internal validation demonstrated excellent discriminative ability, with a C-index of 0.831 (95% CI: 0.780-0.882) for the model and an AUC of 0.838 (95% CI: 0.782-0.893) for the AER index. External validation further confirmed the robustness of both tools, showing no significant difference in predictive performance (p = 0.614).
Conclusion:
The developed model and AER index provide valuable tools for early identification of high-risk SFTS patients, facilitating improved clinical management and resource allocation.