A model for predicting AKI after cardiopulmonary bypass surgery  in Chinese patients with normal preoperative renal function.

Abstract

Objective The objective of this study is to establish and validate a predictive model for the occurrence of acute kidney injury (AKI) following cardiopulmonary bypass (CPB) surgery in Chinese patients with preoperative renal function within normal range. Method From January 2015 to September 2022, a total of 1003 patients were added into the analysis. We used the ratio of 7:3 to divide the patients into a training group (n = 700) and a testing group (n = 303). Independent risk factors for postoperative AKI were identified through the least absolute shrinkage and selection operator (LASSO) regression and multifactor logistic regression analysis. A nomogram predictive model was then established. Various metrics such as the Area Under the ROC curve (AUC), calibration curve, and decision curve were used for validation of the nomogram predictive model in the training and testing groups. Additionally, the nomogram model was compared with three conventional models (Cleveland Clinic score, Mehta score, and Simplified Renal Index (SRI) score) using the AUC, calibration curve, and decision curve. Results The AKI group had a worse prognosis. Age, Body mass index (BMI), emergent surgery, CPB time, intraoperative use of adrenaline, and postoperative procalcitonin (PCT) were identified as important risk factors for AKI after CPB surgery. The nomogram predictive model demonstrated good discrimination (AUC: 0.772 (95%CI: 0.735 − 0.809) and 0.780 (95% CI: 0.724 − 0.835) ), calibration (Hosmer and Lemeshow goodness of fit test: P-value 0.6941 and 0.9539) and clinical utility in both the training and testing groups. Moreover, our model exhibited better discrimination, calibration capacity, and superior net benefit compared to the other three conventional models. Conclusion The nomogram predictive model, which established in patients with normal preoperative renal function, has high accuracy, calibration and clinical utility. The model's performance is superior to the other three conventional models (Cleveland Clinic score, Mehta score, and SRI score).