Abstract
OBJECTIVE: Postoperative venous thromboembolism (VTE) is a potentially life-threatening complication. This study aimed to develop a predictive model to identify independent risk factors and estimate the likelihood of VTE in patients undergoing surgery for cervical cancer. METHODS: We conducted a retrospective cohort study involving 1,174 patients who underwent surgery for cervical carcinoma between 2019 and 2022. The cohort was randomly divided into training and validation sets at 7:3. Univariate and multivariate logistic regression analyses were used to determine the independent factors associated with VTE. The results of the multivariate logistic regression were used to construct a nomogram. The nomogram's performance was assessed via the concordance index (C-index) and calibration curve. Additionally, its clinical utility was assessed through decision curve analysis (DCA). RESULTS: The predictive nomogram model included factors such as age, pathology type, FIGO stage, history of chemotherapy, the neutrophil-lymphocyte ratio (NLR), fibrinogen degradation products (FDP), and D-dimer levels. The model demonstrated robust discriminative power, achieving a C-index of 0.854 (95% CI: 0.799-0.909) in the training cohort and 0.757 (95% CI: 0.657-0.857) in the validation cohort. Furthermore, the nomogram showed excellent calibration and clinical utility, as evidenced by the calibration curve and decision curve analysis (DCA) results. CONCLUSIONS: We developed a high-performance nomogram that accurately predicts the risk of VTE in cervical cancer patients undergoing surgery, providing valuable guidance for thromboprophylaxis decision-making.