Abstract:

Objective To review the related studies on the risk prediction model of extracorporeal circulation coagulation in patients with continuous renal replacement therapy. Methods The searches of Chinese and English databases were conducted to collect relevant literature on predictive models of coagulation risk in extracorporeal circulation in patients undergoing continuous renal replacement therapy that were publicly available from the time of database construction to November 8，2024，and the descriptive statistical analysis was conducted. Results A total of 8 literatures，mostly single-center studies，were included，with a modeling sample size of 79 to 1，752 cases and the number of outcome events ranging from 51 to 558. The model construction methods included cox proportional hazard model，logistic regression model and machine learning，and the area under the receiver operating characteristic curve was 0.732~0.898. The predictors that appeared the most frequently were anticoagulation modality，blood flow rate，APTT and platelet count. Conclusion The risk prediction model of extracorporeal circulation coagulation in continuous renal replacement therapy is of great significance for identifying high-risk patients in advance，guiding clinical decision-making，reducing the incidence of coagulation and improving patient prognosis. The current models had good predictive efficiency，but the model evaluation and validation were insufficient，and the overall risk of bias was high. Future studies need to optimize existing models and develop more stable and clinically applicable predictive models.

Key words: Continuous Renal Replacement Therapy, Extracorporeal Circulation, Blood Coagulation, Models, Statistical, Scoping Review