Derivation and validation of a nomogram model for venous thromboembolism in patients undergoing borderline resectable pancreatic cancer surgery

Background: Venous thromboembolism (VTE) is a common postoperative complication in patients with borderline resectable pancreatic cancer (BRPC), which seriously affects the quality of life and prognosis. We aimed to establish new nomogram to predict the VTE risk after BRPC surgery. Method: A total of 301 consecutive adult patients who underwent BRPC surgery between Jan. 2011 and Dec. 2021 in our institution were retrospectively reviewed. Included in the cohort were 67 patients who developed VTE and 79 non-VTE patients. The patients were randomly divided into a training set and a validation set at a ratio of 7:3. A nomogram model was developed based on the results of univariate and multivariate logistic regression analysis in the training group. The predictive power of the nomogram was evaluated by receiver operating characteristic curve (ROC), calibration curve, and decision curve analysis (DCA). Results: : The incidence of VTE in the present study was 45.9%. Age, albumin, operation time, tumor diameter and preoperative D-dimer were independent related factors. The areas under the ROC curve of the nomogram were 0.843 (95% CI: 0.782-0.904) and 0.893 (95% CI: 0.834-0.953) in the training and validation groups, respectively. For clinical usefulness, the model offered a net benefit in decision curve analysis, suggesting the new nomogram predict VTE events accurately enough. Conclusion: We first established a high-performance nomogram of VTE risk in patients undergoing BRPC surgery, which helps clinicians to screen out people at high risk of VTE and perform prophylactic anticoagulation.