Analysis of the predictive value of Caprini score after colorectal cancer surgery and development of a new risk assessment model: A single-centre prospective cohort study in China

Background: This study aimed to assess the Caprini score in predicting venous thromboembolism (VTE) events and develop a risk assessment model to predict the occurrence of short-term VTE after colorectal cancer surgery. Methods: The study included 181 patients who underwent colorectal cancer surgery at Beijing Friendship Hospital, from October 2020 to May 2022. The results of the Caprini score 24 h after surgery were recorded to evaluate its predictive value. Multifactorial staging was used to determine independent risk factors for VTE, and nomogram models were constructed based on these factors. Sixty-one patients at Beijing Chaoyang Hospital were used for external validation. Results: The incidence of VTE after colorectal cancer was 17.6% in 32 patients. Age (odds ratio [OR] 1.106, 95% confidence interval [CI] 1.049–1.167, P < 0.001), sex (female) (OR 2.981, 95% CI 1.176–7.557, P = 0.021), and cardiovascular disease [CVD] (OR 5.611, 95% CI 2.043–15.413, P = 0.001) were independent risk factors for postoperative VTE. The nomogram model showed good predictive performance (area under the curve [AUC], 0.837 [95% CI, 0.762–0.917]; Nagelkerke R2, 0.347; Brier score, 0.106) compared with the Caprini score (AUC, 0.760 [95% CI, 0.674–0.847]; Nagelkerke R2, 0.104; Brier score, 0.131). The decision curve analysis showed that the nomogram model had a better net clinical benefit. Conclusions: The Caprini score is moderately efficient in predicting the occurrence of VTE after colorectal cancer surgery, and the 3-item nomogram model can be used as a predictive tool in VTE prevention decisions after colorectal cancer surgery.