A Novel Predictive Model for Poor In-hospital Outcomes in patients with acute kidney injury after cardiac surgery.

Abstract: Objective Patients with cardiac surgery-associated acute kidney injury (CSA-AKI) are at risk of renal replacement therapy (RRT) and in-hospital death. We aimed to develop and validate a novel predictive model for poor in-hospital outcomes among CSA-AKI patients. Methods A total of 196 patients diagnosed with CSA-AKI were enrolled in this study as the training cohort, and 32 blood cytokines were measured. Least absolute shrinkage and selection operator (LASSO) regression and random forest quantile-classifier (RFQ) were performed to identify the key blood predictors for in-hospital composite outcomes (requiring RRT or (and) in-hospital death). The logistic regression model incorporating the selected predictors was validated internally using bootstrapping and externally in an independent cohort (n=52). Results A change in serum creatinine (delta sCr) and interleukin (IL)-16 and IL-8 were selected as key predictors for composite outcomes. The logistic regression model incorporating IL-16, IL-8 and delta sCr yielded the optimal performance, with decent discrimination [area under the receiver operating characteristic curve (ROC-AUC): 0.947; area under the precision-recall curve (PR-AUC): 0.809] and excellent calibration (Brier score: 0.056, Hosmer–Lemeshow test p =0.651). Application of the model in the validation cohort still yielded good discrimination. A nomogram was generated for clinical use, and decision curve analysis (DCA) demonstrated that the new model adds more net benefit than delta sCr. Conclusions We developed and validated a promising predictive model for in-hospital composite outcomes among CSA-AKI patients and demonstrated IL-16 and IL-8 as useful predictors to improve risk stratification for poor in-hospital outcomes among CSA-AKI patients.