RESUMO
BACKGROUND AND OBJECTIVES: Lung cancer patients slated for surgery are at high risk of venous thromboembolism (VTE). Precise risk assessment is necessary for providing proper thromboprophylaxis and reducing morbidity and mortality of VTE. METHODS: A multicenter, observational, cross-sectional cohort study, involving patients with primary lung cancer undergoing surgery, was carried out from August 2016 to December 2019. All patients were assessed according to the Caprini risk assessment model (RAM) and a modified scoring system incorporating elevated D-dimer and new stratification of surgical time. The endpoint was confirmed VTE or patient discharge. RESULTS: Out of 1205 patients, 87 (7.2%) were diagnosed with VTE. The area under the curve of modified scores for VTE was 0.759, which was larger than that of the original one (0.589) (p < 0.05). By modified Caprini scoring system, a higher score was associated with increased VTE risk (odds ratio [OR], 1.345; 95% confidence interval [CI], 1.197-1.512; p < 0.001), and there was an increased OR of 4.090 (95% CI, 2.472-6.768, p < 0.001) for VTE in high-risk category patients. CONCLUSION: Modified Caprini RAM showed an improved prediction of high-risk patients with an elevated likelihood of postoperative VTE compared to the original one.