Clinical Prediction Models for Preoperative Acute Kidney Injury due to Acute Aortic Dissection with Different Typing
Objective:To analyze the clinical data of patients with acute aortic dissection(AAD)in a single center in recent 3 years,and to construct clinical prediction models of acute renal injury in patients with different types of AAD,in order to guide relevant clinicians in early diagnosis and prevention of AKI.Methods:The clinical data of patients hospitalized and diagnosed with acute aortic dissection at the First Affiliated Hospital of Xinjiang Medical University from January 2019 to December 2021 were retrospectively col-lected,and they were classified into type A and type B patients according to CTA imaging findings,and further divided into AKI and non-AKI groups,respectively,according to KDIGO criteria.The differences in preoperative clinical data between the two groups were compared,and the independent predictors of AAD-AKI were screened by binary logistic regression,and a logistic clinical prediction model was constructed.The ROC curves of each independent predictor and the combined diagnosis were plotted to evaluate the diagnos-tic value of the prediction model for the preoperative occurrence of AKI.Results:Multivariate Logistic regression analysis showed that systolic blood pressure,white blood cells and first-time creatinine were independent predictors of AKI in patients with TAAAD.TAAAD-AKI prediction model Logit(P)=-5.189-0.019*systolic blood pressure+0.109*leukocytes+0.012*creatinine for the first time on admission.The independent predictors of TBAAD patients were lactic acid concentration,creatinine for the first time on admis-sion and poor renal perfusion,and the predictive model was Logit(P)=-2.976-0.295*lactic acid concentration+0.042*creatinine for the first time on admission+0.655*renal poor perfusion.Conclusion:The best univariate predictor of AKI in patients with type An and type B is creatinine for the first time on admission,and the mechanism of AKI in patients with different types of dissection is not the same.Therefore,according to the above predictors,we have established clinical models and found that they all have good predictive titers,which can provide a basis for clinicians'diagnosis.
Acute aortic dissectionAcute renal injuryPredictors and prediction models
NETL
NSTL
万方数据
