Abstract
OBJECTIVE: To develop and evaluate a nomogram model for predicting prolonged ICU length of stay (LOS) in patients with sepsis-induced coagulopathy (SIC), identify associated risk factors, and facilitate early identification of high-risk patients, with the aim of optimizing clinical management strategies, improving patient outcomes, and enhancing ICU resource utilization. METHOD: A total of 3728 patients meeting the diagnostic criteria of International Society for Thrombosis and Hemostasis (ISTH) criteria were included from the Medical Information Mart for Intensive Care (MIMIC-IV) database. Based on the third quartile value of ICU LOS in the cohort, patients were categorized into a prolonged ICU LOS group (≥ 5 days) and a non-prolonged ICU LOS group (< 5 days). General demographic data, clinical characteristics, and laboratory test results within 24 h of ICU admission were collected to identify independent risk factors for prolonged ICU LOS in SIC patients. Predictive variables were selected using LASSO-logistic regression combined with Shapley Additive Explanations (SHAP) for interpretability. A nomogram model was constructed and evaluated using receiver operating characteristic (ROC) curves, calibration curves, and decision curve analysis (DCA). RESULTS: Among the 3728 enrolled patients, 832 had a prolonged ICU LOS (≥ 5 days), while 2896 had a non-prolonged ICU LOS (< 5 days). LASSO-logistic regression and SHAP analysis identified six predictive variables: SOFA score, heart rate, monocyte percentage, acute kidney injury (AKI), use of vasopressors, and use of mechanical ventilation. The nomogram demonstrated an area under the curve (AUC) of 0.737 (95% CI 0.716-0.758). CONCLUSION: Risk factors for prolonged LOS in patients with SIC included: increased SOFA score, elevated heart rate, higher monocyte percentage, occurrence of AKI, use of vasopressors, and use of mechanical ventilation. By integrating these readily available clinical indicators into an intuitive nomogram, we have developed a practical risk assessment tool for clinicians. This tool aids in the identification of SIC patients at risk for prolonged hospitalization. Furthermore, our analysis revealed that prolonged LOS was significantly associated with increased long-term mortality. The application of this predictive model may ultimately contribute to reducing ICU length of stay and improving patient prognosis.