To develop and validate a nomogram model to predict postoperative 30-day mortality in acute type A aortic dissection patients receiving total aortic arch replacement with frozen elephant trunk technique.
Clinical data on 1,156 consecutive acute type A aortic dissection patients who got total aortic arch replacement using the frozen elephant trunk technique was collected from January 2010 to December 2020. These patients were divided into training and testing cohorts at random with a ratio of 7:3. To predict postoperative 30-day mortality, a nomogram was established in the training set using the logistic regression model. The novel nomogram was then validated in the testing set. The nomogram's calibration and discrimination were evaluated. In addition, we created four machine learning prediction models in the training set. In terms of calibration and discrimination, the nomogram was compared to these machine learning models in testing set.
Left ventricular end-diastolic diameter <45 mm, estimated glomerular filtration rate <50 ml/min/1.73 m2, persistent abdominal pain, radiological celiac trunk malperfusion, concomitant coronary artery bypass grafting and cardiopulmonary bypass time >4 h were independent predictors of the 30-day mortality. The nomogram based on these 6 predictors manifested satisfying calibration and discrimination. In testing set, the nomogram outperformed the other 4 machine learning models.
The novel nomogram is a simple and effective tool to predict 30-day mortality rate for acute type A aortic dissection patients undergoing total aortic arch replacement with frozen elephant trunk technique.