Skip to content
Open access

Development and Validation of an Interpretable Machine Learning Model to Predict Mortality in Patients With Sepsis-Induced Coagulopathy: Multicenter Cohort Study

Sep 2026 · JMIR Medical Informatics · Vol 14 · 0 citations · 50 references
Medicine

TL;DR

The ML model developed in this study demonstrated superior accuracy, robustness, and generalizability in predicting early mortality in patients with SIC, and the findings provide valuable insights for prognostic assessment and the individualized management of patients with SIC.

Abstract

Abstract Background Sepsis-induced coagulopathy (SIC) is a common and severe complication in patients with sepsis, characterized by microvascular thrombosis, systemic endothelial damage, and markedly increased short-term mortality. Existing traditional clinical risk scoring systems demonstrate limited accuracy and fail to capture complex, nonlinear physiological interactions, underscoring the urgent need for advanced prognostic tools. Objective The objective of this study was to develop and validate an interpretable machine learning (ML) model using large-scale, multicenter databases to predict early mortality in intensive care unit (ICU) patients with SIC and to evaluate its predictive performance and clinical utility compared with traditional clinical risk scores. Methods The study retrospectively analyzed clinical data of patients with SIC from the Medical Information Mart for Intensive Care IV (MIMIC-IV), the eICU Collaborative Research Database (eICU-CRD), and Tianjin Medical University General Hospital. Feature selection was performed using LASSO (least absolute shrinkage and selection operator) regression, the Boruta algorithm, and recursive feature elimination with cross-validation, combined with multivariable logistic regression. Twelve ML algorithms were trained and compared with 4 traditional clinical scoring systems (Sequential Organ Failure Assessment, Acute Physiology and Chronic Health Evaluation II, Simplified Acute Physiology Score II, and Oxford Acute Severity of Illness Score) to predict 28-day mortality after ICU admission. Model discrimination, calibration, and clinical utility were assessed using the area under the curve (AUC), calibration curves, and decision curve analysis. Furthermore, Shapley additive explanations (SHAP) values were used to ensure model interpretability and to identify individual pathophysiological drivers. Results A total of 7980 (66.3%) patients with SIC from the MIMIC-IV database, 3815 (31.7%) from the eICU-CRD database, and 235 (2.0%) from Tianjin Medical University General Hospital were included in the study. Ten independent predictors were identified to construct the model. The XGBoost (extreme gradient boosting) model performed the best, with an AUC of 0.899 (95% CI 0.889‐0.909) in the internal validation set and 0.882 and 0.904 in the 2 external validation sets, significantly outperforming traditional scoring systems and demonstrating higher clinical net benefit. SHAP analysis identified that the top 5 critical features were anion gap, red blood cell distribution width, lactate, total bilirubin, and age. Furthermore, an easy-to-use online tool, SIC Predict Streamlit, was developed based on this model to enable clinicians to quickly assess patient risk. Conclusions The ML model developed in this study demonstrated superior accuracy, robustness, and generalizability in predicting early mortality in patients with SIC. The model outperformed traditional clinical scoring tools, and the findings provide valuable insights for prognostic assessment and the individualized management of patients with SIC.

Read PDF

Similar papers

Open access Sep 2026

Development of an interpretable machine learning model for predicting in-hospital mortality in ICU patients with sepsis: a retrospective cohort study

An interpretable RF-based machine learning model for predicting in-hospital mortality in patients with sepsis was developed and internally validated and showed acceptable discrimination, reasonable calibration, and potential clinical net benefit in internal validation.

Tian-Yu Zhao, Ke-Xin Wen, Xu-Min Han et al. · 0 citations
Open access Oct 2026

Development and validation of an interpretable machine learning model for early prediction of mortality in ICU patients with chronic obstructive pulmonary disease and heart failure

Chronic obstructive pulmonary disease and heart failure frequently coexist, and the mortality risk for patients with these comorbidities is significantly elevated in the intensive care unit (ICU) setting. Therefore, developing an interpretable model for early mortality prediction is crucial to improving the prognosis a...

Jian-Ying Shen, Yu-Cheng Liu, Yi-Kun Chen · 0 citations
Open access Oct 2026

Development, Validation, and Deployment of an Interpretable Machine Learning Model for In-Hospital Mortality Prediction in Critically Ill Patients with COPD and Sepsis: A Multicenter Cohort Study

Background Critically ill patients with chronic obstructive pulmonary disease (COPD) and sepsis are at high risk of in-hospital mortality because of respiratory and organ dysfunction. Early risk assessment may be useful for clinical decision-making. This study aimed to develop an interpretable machine learning model fo...

Jian-Jie Ju, Jing-Jing Chen, Gao-Jin Li et al. · 0 citations
Open access Sep 2026

Development and Validation of an Interpretable Machine Learning-Based Prediction Model of Sepsis: A Retrospective Multicenter Cohort Study.

Although the model shows potential for identifying high-risk individuals at the 24-hour prediction landmark, further prospective validation and clinical impact assessment are required before routine clinical implementation.

Zhang-Cong Lu, Ling-Han Hu, Peng-Fei Shui et al. · 0 citations
Open access Oct 2026

An interpretable machine learning model using routine clinical indicators for early pediatric sepsis risk stratification: a multicenter retrospective cohort study.

Background Sepsis remains a leading cause of pediatric morbidity and mortality worldwide. The 2024 Phoenix Sepsis Criteria emphasize early recognition of organ dysfunction. Conventional diagnostics like blood cultures are time-consuming, and existing machine learning models lack early pathogen-level classification. A m...

Ya-Dong Li, X. Li, Dong Wang et al. · 0 citations
Open access Oct 2026

Development and validation of a prognostic prediction model for elderly patients with sepsis-associated acute kidney injury based on interpretable machine learning

Background Elderly patients with sepsis-associated acute kidney injury (SA-AKI) have high short-term mortality; early identification of high-risk individuals is critical for personalized intervention. Traditional clinical scoring systems relying on linear assumptions and conventional machine learning models often lack...

Nan Zhang, Jian-Zhong Wang, Qi Ma et al. · 0 citations

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.