Development and validation of a nomogram for early prediction of sepsis-induced coagulopathy: a multicenter study

Sepsis-induced coagulopathy (SIC) is a vascular endothelial cell injury and coagulation disorder caused by sepsis. The aim of this study was to construct a nomogram model of the risk of early onset of SIC in patients with sepsis by analyzing the risk factors for in-hospital development of SIC.

Patients with sepsis admitted to the intensive care unit (ICU) of Hebei General Hospital and Handan Central Hospital (East District) from March 1, 2021 to March 1, 2024 were retrospectively included. Sepsis patients were divided into SIC and non-SIC groups according to whether SIC occurred during hospitalization. The patient data were randomly divided into training set and testing set in the ratio of 7:3. The data of sepsis patients admitted to the ICU of Hebei General Hospital between March 1, 2024 and October 31, 2024 were then retrospectively included as the validation set for external validation. All predictors were collected within 24 h of sepsis diagnosis to enable early risk prediction. Various clinical variables were collected, and independent risk factors for early onset of SIC were screened by one-way logistic regression, least absolute shrinkage and selection operator (LASSO) regression, and multifactorial logistic and a nomogram prediction model was constructed. The model was evaluated for accuracy, goodness of fit, and clinical utility value using testing set and validation set data. The accuracy of the predictive model was assessed by using the receiver operating characteristic curve (ROC) and calculating the area under the receiver (AUC), the fit was done by calibration curve, and the clinical utility of the predictive model was assessed by decision curve analysis (DCA).

Among 847 patients with sepsis, SIC occurred in 480 (56.7%) patients. A nomogram model was constructed containing eight variables: lactate, oxygenation index, total protein, total bilirubin, urea, calcitoninogen, activated partial thromboplastin time, and monocyte count. In the training set, the AUC value of the model was 0.783 [95% Confidence Interval (CI): 0.746, 0.820]; in the testing set, the AUC value was 0.768 (95% CI: 0.710, 0.826); and in the validation set, the AUC value was 0.782 (95% CI: 0.708, 0.856).

We developed a nomogram model to predict the risk of SIC in patients with sepsis and validated its potential as a clinically reliable tool. The overall accuracy and clinical utility value of the model was high and the fit was good. The nomogram model can visualize the key variables associated with SIC in sepsis patients, supporting clinicians in individualized risk assessment and guiding timely interventions to improve patient outcomes.