Abstract
Background: The risk factors and association of venous thromboembolism (VTE) following emergent ventral hernia repair (EVHR) remains uncertain. This aim of the study aims was to establish the predictors of VTE after EVHR and its influence on the long-term outcomes. Methods: A total of 2093 patients from the MIMIC-IV database who underwent EVHR were recruited. Multivariate logistic regression and nomogram models were developed to predict in-hospital VTE and mortality. Calibration and receiver operating characteristic (ROC) curves were utilized to assess the model's effectiveness and reliability. Decision curve analysis (DCA) was performed to evaluate the net clinical benefits of the model. Results: The rate of in-hospital VTE was 1.6% (33/2093) after EVHR. Four independent potential factors were established after multivariate analysis, and the abovementioned risk factors fit into the nomogram. The prediction model presented good performance metrics (C-index: 0.857), the calibration and ROC curves demonstrated the accurate prediction power, and DCA indicated the superior net benefit of the established model. In-hospital and 1-year mortality rates were 0.8% (17/2093) and 4.1% (86/2076) after EVHR. The potential factors were included in the mortality prediction nomogram. The prediction model presented good performance metrics (C-index of 0.957 and 0.828, respectively), the calibration and ROC curves were consistent with the actual results, and DCA indicated the superior net benefit of the established model. Conclusion: The nomogram, derived from the logistic regression model, demonstrated excellent predictive performance for VTE occurrence and prognosis in patients following EVHR. This model could serve as a valuable reference for clinical decision-making regarding VTE prevention and for enhancing post-EVHR prognosis.