Abstract: Objective  This study aimed to investigate the factors influencing extracorporeal circulation failure during continuous renal replacement therapy (CRRT) in pediatric patients and to develop and validate a predictive model.  Methods  A retrospective analysis was conducted on data from 370 pediatric patients who received CRRT at Guangdong Provincial People's Hospital between 2014 and 2021. Patients were divided into failure and non-failure groups based on the occurrence of extracorporeal circulation failure within 24 hours of CRRT initiation. Through the application of the Lasso regression regularization method and clinical judgement, the key variables associated with extracorporeal circulatory failure were identified and selected. These variables were then incorporated into a multivariate logistic regression analysis to identify independent risk factors, leading to the construction of a predictive model.  Results  The final predictive model was based on body weight (OR=0.932, 95% CI: 0.910～0.954, P＜0.001), mean ultrafiltration rate (OR=1.012, 95% CI: 1.007～1.017, P＜0.001), insufficient blood flow (OR=26.535, 95% CI: 12.507～56.297, P＜0.001), and lactate levels (OR=1.095, 95% CI: 1.014～1.183, P=0.021). The area under the receiver operating characteristic (ROC) curve for the model was 0.891 (85.61%～92.48%). Calibration curves and clinical decision curves indicated that the model has good predictive capability and clinical utility.  Conclusion  Low body weight, high mean ultrafiltration rate, insufficient blood flow, and elevated lactate levels are independent risk factors for extracorporeal circulation failure within 24 hours of CRRT in pediatric patients. The predictive model based on these factors demonstrates high predictive accuracy.

Key words: Pediatrics, Continuous renal replacement therapy, Extracorporeal circulation obstruction, Predictive model