A nomogram for reduced cardiac function in postoperative acute type A aortic dissection patients with acute kidney injury undergoing continuous renal replacement therapy

Abstract

Background: This study aimed to develop a nomogram to predict reduced cardiac function for acute kidney injury (AKI) patients who received continuous renal replacement therapy (CRRT) after acute type A aortic dissection (ATAAD) surgery.

Methods: This study was a retrospective analysis. ATAAD patients with preoperative normal ejection fraction (EF) and postoperative AKI with CRRT admitted between January 2014 and November 2021 were included. The reduced cardiac function was defined as EF <50%. The data were analyzed by the univariate and multivariate logistic regression analyses. A diagnostic model was established by a nomogram, and its discriminative performance was validated by the received operating characteristic (ROC) curve and concordance (C) statistic. The calibration of the diagnostic model was tested by calibration curves and the HosmerLemeshow test. The clinical utility was evaluated by the decision curve analysis (DCA).

Result: In total, 208 patients were eligible for analysis, of which 98 patients with reduced cardiac function. The logistic regression analyses showed age ≥60 years old, history of coronary atherosclerotic disease, preoperative pericardial tamponade, and cardiopulmonary bypass time were risk factors for reduced cardiac function, which were further employed in the nomogram. As results, nomogram revealed a high predictive power (C statistic = 0.723, 0.654-0.792; the bootstrap-corrected concordance C statistic = 0.711, the area under the ROC curve = 0.723). The calibration curves showed good consistency between the predicted and the actual probabilities (calibration curve: Brier points = 0.208, Emax = 0.103, Eavg = 0.021; Hosmer-Lemeshow test, P = 0.476). DCA showed that the nomogram could augment net benefits and exhibited a wide range of threshold probabilities in the prediction of EF reduction.

Conclusion: This nomogram is an effective diagnostic model for predicting the reduced cardiac function in postoperative ATAAD patients with AKI undergoing CRRT and can be used to protect postoperative renal functions and facilitate patient-specific care after ATAAD surgery.

Keywords: acute kidney injury; acute type A aortic dissection; continuous renal replacement therapy; diagnostic model; postoperative ejection fraction reduction.

Figure 1

Supervised tree-like segmentation of EF.

Figure 2

The study flowchart.

Figure 3

Nomogram predicts EF reduction risk in patients with AKI undergoing CRRT after ATAAD surgery. The nomogram was established to predict the risk of decreased EF in patients with CRRT based on 4 independent prognostic factors. The value of each of variable wasgiven a score on the point scale axis. The total score can be calculated by summation of single scores. We can estimate the probability of decreased EF by projecting the total score to the lower total point scale.

Figure 4

Receiver operating characteristic (ROC) curve for evaluating the discrimination performance of the model, area under ROC curve was 0.723, and concordance (C) statistic was 0.723.

Figure 5

Calibration curves for the prediction model. The curves describe the calibration of the nomogram in terms of the agreement between predicted risks (X-axes) and actual outcomes (Y-axis). The diagonal line indicates a perfect prediction by an ideal model. E max, maximum error; E aver, average error.

Figure 6

The decision curve analysis (DCA) for the prediction model. The DCA showed that the nomogram could augment net benefits and exhibited a wide range of threshold probabilities in the EF reduction.