Development and validation of a nomogram for postoperative severe acute kidney injury in acute type A aortic dissection

BACKGROUND Postoperative acute kidney injury(AKI) is a major complication associated with increased morbidity and mortality after surgery for acute type A aortic dissection(AAAD). To the best of our knowledge, risk prediction models for AKI following AAAD surgery have not been reported. The goal of the present study was to develop a prediction model to predict severe AKI after AAAD surgery.METHODS A total of 485 patients who underwent AAAD surgery were enrolled and randomly divided into the training cohort(70%) and the validation cohort(30%). Severe AKI was defined as AKI stage Ⅲ following the Kidney Disease: Improving Global Outcomes criteria. Preoperative variables, intraoperative variables and postoperative data were collected for analysis. Multivariable logistic regression analysis was performed to select predictors and develop a nomogram in the study cohort. The final prediction model was validated using the bootstrapping techniques and in the validation cohort.RESULTS The incidence of severe AKI was 23.0%(n = 78), and 14.7%(n = 50) of patients needed renal replacement treatment.The hospital mortality rate was 8.3%(n = 28), while for AKI patients, the mortality rate was 13.1%, which increased to 20.5% for severe AKI patients. Univariate and multivariate analyses showed that age, cardiopulmonary bypass time, serum creatinine, and D-dimer were key predictors for severe AKI following AAAD surgery. The logistic regression model incorporated these predictors to develop a nomogram for predicting severe AKI after AAAD surgery. The nomogram showed optimal discrimination ability, with an area under the curve of 0.716 in the training cohort and 0.739 in the validation cohort. Calibration curve analysis demonstrated good correlations in both the training cohort and the validation cohort.CONCLUSIONS We developed a prognostic model including age, cardiopulmonary bypass time, serum creatinine, and D-dimer to predict severe AKI after AAAD surgery. The prognostic model demonstrated an effective predictive capability for severe AKI, which may help improve risk stratification for poor in-hospital outcomes after AAAD surgery.