Development of a nomogram to predict in-hospital mortality of lower extremity deep vein thrombosis patients in intensive care unit

Abstract

Object: This study aimed to analyze factors influencing the survival of patients with lower extremity deep vein thrombosis (LEDVT) complicated during intensive care unit (ICU) admission and develop a nomogram for predictive use. Materials and Methods: Retrospective analysis of 482 LEDVT cases in ICU at Third Xiangya Hospital (Feb 2017-Nov 2022). Data on demographics, comorbidities, treatments, vital signs, and lab indicators (within 24 hours of admission) were collected. Patients were randomly divided into training and test sets (3:1 ratio). Univariate and multivariate analyses were performed using SPSS and RStudio. Logistic regression and nomogram models were constructed and validated. Results: Intubation/Tracheotomy, vasopressor usage, age, lymphocytes and lactate were identified as independent risk factors. rivaroxaban usage and enoxaparin usage were identified as protective factors. The nomogram had area under the curve(AUC) of 0.855 (training set) and 0.775 (testing set). Decision curve and clinical impact analyses demonstrated the nomogram's clinical utility within specific probability ranges. Conclusion: The developed nomogram accurately predicts in-hospital mortality of LEDVT patients in the ICU, providing valuable insights for clinical decision-making and patient management.