Predictive Factors for In-Hospital Preoperative Rupture in Hyperacute Type A Aortic Dissection.

ABSTRACT

OBJECTIVE: This study aims to figure out risk factors of in-hospital preoperative rupture of hyperacute type A aortic dissection (haTAAD) patients and build a prediction and risk stratification model. METHODS: From January 2011 to December 2019, 830 patients diagnosed as haTAAD from Nanjing Drum Tower Hospital were enrolled. Among them, 799 patients received prompt surgery and 31 suffered aortic rupture before operation. The association between in-hospital preoperative rupture and perioperative parameters were examined. Best subset selection was used for feature selection and ROC curve was used to identify the model. RESULTS: Age, winter season, back pain, preoperative hypotension, albumin and globulin ratio, high serum phosphorus level are risk factors for in-hospital preoperative rupture of haTAAD. On the basis of six variables with AUC 0.828, a nomogram was established. According to the robustness test, actual in-hospital preoperative ruptures were fitted well. CONCLUSIONS: The in-hospital rupture prediction model was developed using logistic regression analysis. High serum phosphorus level is one of the strongest predictors. This nomogram may be useful when evaluating the risk of aortic dissection in-hospital rupture in future trials.