Abstract
OBJECTIVE: To develop and validate a risk prediction model for postoperative venous thromboembolism (VTE) in elderly patients (≥ 60 years old) with lung cancer using internal and external datasets, providing a reference for clinical prevention and management. METHODS: A retrospective analysis was conducted on 320 elderly lung cancer patients who underwent surgery at our hospital between January 2023 and May 2024. Patients were categorized into a VTE group (55 cases) and a non-VTE group (265 cases) based on the occurrence of VTE within six months postoperatively. An additional 60 elderly lung cancer patients who underwent surgery between June and November 2024 were selected for external validation. General clinical characteristics were compared between the two groups, and risk factors for postoperative VTE were analyzed to construct and validate the predictive model. RESULTS: The incidence of postoperative VTE was 17.19% (55/320). Compared with the non-VTE group, the VTE group had a significantly higher proportion of patients with a history of smoking, alcohol consumption, hypertension, extremely high-risk Caprini scores, surgical duration ≥ 120 min, open surgery, clinical stage III-IV disease, preoperative chemotherapy, positive D-dimer (D-D) >3.64 mg/L, and elevated C-reactive protein (CRP) levels (P < 0.05). Multivariate analysis identified extremely high-risk Caprini scores, surgical duration ≥ 120 min, open surgery, clinical stage III-IV disease, preoperative chemotherapy, positive D-D>3.64 mg/L, and elevated CRP levels as independent risk factors for postoperative VTE (P < 0.05). Receiver operating characteristic (ROC) curves, calibration curves, and decision curve analysis (DCA) demonstrated that the model had excellent discrimination, accuracy, and predictive performance. CONCLUSION: The incidence of postoperative VTE in elderly lung cancer patients is relatively high. Key risk factors include an extremely high-risk Caprini score, surgical duration ≥ 120 min, open surgery, clinical stage III-IV disease, preoperative chemotherapy, positive D-D>3.64 mg/L, and elevated CRP levels. The nomogram-based risk prediction model developed in this study exhibits high discrimination and accuracy, with good predictive ability and clinical utility, providing a basis for targeted prevention and management strategies.