There was a lack of studies that have explored the prognostic value of AAD for postoperative AKI in cardiac surgery[3, 16, 18]. This is the first study to explore the prognostic value of AAD for AKI. In this study, we retrospectively collected the preoperative and intraoperative data, minimized the effect of the confounding factors through various methods, and finally clarified the prognostic value of AAD for AKI after cardiac surgery in adult patients, especially for on-pump surgeries. The main finding of this study is that patients with a higher AAD preoperatively tend to have a higher risk of AKI after cardiac surgery. In addition, in females, lower albumin, longer CPB time, higher ascending aorta diameter, higher left atrial anteroposterior diameter, higher left pulmonary artery diameter, and lower right pulmonary artery diameter are also independent risk factors.
In this study, the incidence of postoperative acute kidney injury (AKI) was 10.41%, consistent with the range reported in previous studies[19]. The results suggested that in cardiac surgery patients, both univariate and multivariate analyses showed significant statistical significance for preoperative ascending aortic diameter (AAD), indicating its predictive value for postoperative AKI occurrence.
In the overall 442-patient cohort, multivariate logistic regression identified that AAD was a prognostic factor for postoperative AKI. The results showed that gender, serum albumin, and CPB time were risk factors for postoperative AKI after cardiac surgery, which was consistent with previous studies. However, this study focused mainly on the predictive value of cardiac ultrasound findings and did not include all relevant factors. Further studies are needed to explore other factors.
In all patients, the prognostic value of AAD showed statistically significant differences based on the univariate and multivariate analyses. However, the ROC value of AAD for AKI was 0.61, indicating the limited prognostic value of AAD in all cohort patients. Therefore, we did not explore the cut-off value further. Additionally, there was no statistically significant difference between models with and without AAD.
Patients who underwent CPB had a higher risk of postoperative AKI, and the CPB application might be a confounding bias in statistical analysis. Subgroup analysis in on-pump patients showed that the AAD had prognostic value based on both univariate and multivariate analysis results (P < 0.05), and improved relevant models (P < 0.05). AAD had a higher AUC value than that of CPB in the on-pump patients subgroup analysis, indicating its relatively better prognostic value. Further studies are needed to explore the effect of other confounding factors on AKI after cardiac surgery.
This study found that in adult cardiac surgery patients, higher preoperative AAD values were associated with a higher risk of postoperative AKI. The difference in AAD average between the two groups was below 3mm clinically but it showed great significance[20]. This can be attributed to several reasons: Firstly, the ascending aorta is the major artery and supplies oxygen and perfusion of the kidney[6]. Secondly, the kidney's arterial branches are derived from the ascending aorta. Therefore, its dilation reflects renal vascular function and compensatory capacity in some way[15]. Thirdly, patients undergoing CPB are at higher risk of postoperative AKI. The preoperative conditions of the ascending aortic such as dilation or obstruction might increase relative risks through changed pathophysiology[21]. Therefore, AAD measurement before surgery should be paid great importance, especially for these high-risk patients.
The advantages of this study are as follows. Firstly, the sample size was enough to prove the evidence based on the statistical calculation. Secondly, we included multiple potential factors affecting postoperative AKI and used univariate, multivariate, ROC, and subgroup analysis statistical methods to control the confounding factors.
However, there are some limitations of this study. First, the design of this study is a retrospective study. Although we used the imputation statistical method, the bias of retrospective studies should be rectified by further prospective studies or larger sample size studies. Second, this study included limited confounding factors, further studies are needed to include more variables. Thirdly, it is unclear the cut-off value of AAD, further studies are encouraged to find a reasonable cut-off value.