Construction of risk prediction model for acute type A aortic dissection

Objective To analyze the risk factors associated with acute type A aortic dissection, and to establish an effective predictive scoring model for early clinical intervention. Methods A retrospective review of 200 patients diagnosed as acute type A aortic dissection in the Department of Emergency, Guangdong Provincial People's Hospital between April 1, 2012 and March 31, 2017 was conducted as a modeling group. There were 160 males and 40 females with age of (53.30 ±13.19) years. According to the criteria of dissection rupture within 72 h after CTA examination, patients were divided into a rupture group (100 cases) and an unruptured group (100 cases). A prediction model of rupture risk was established based on the multivariate Logistic regression analysis of the risk factors of acute A-type aortic dissection in the early stage. After the establishment of the model, a prospective review of 80 patients diagnosed as acute type A aortic dissection in our hospital between April 1, 2017 and May 31, 2018 was conducted as a validation group. There were 57 males and 23 females with age of (54.00 ±13.68) years. In the validation group, there were 53 patients with rupture and 27 patients without rupture. Furthermore, this model was verified in the validation group. The scores were divided into low-risk rupture and high-risk rupture according to the best cut-off value of the receiver operating model curve of the modeling group. Results Logistic regression analysis showed that 10 factors were finally established the rupture risk prediction score model. The factors were as following: age >63 years (2 points), women (2 points), ventilator assisted ventilation (3 points), AST >80 U/L (2 points), no distortion of the inner membrane (2 points), diameter of the aortic sinus >41 mm (1 point), maximum diameter of the ascending aorta >48 mm (1 point), ratio of false lumen area to true lumen area >2.12 (2 points), Lac >1.9 mmol/L (3 points), and WBC >14.2×109 /L (1 point). The results of validation showed that the scoring model had a high predictive value (AUC = 0.928, 95% CI: 0.872-0.984, P < 0.001) and goodness of fit (Hosmer-Lemeshow χ2 = 8.331, P = 0.402). Moreover, the scoring model was further divided into low-risk (0-6 points) and high-risk (7-19 points), with sensitivity of 83.0% and specificity of 86.0%. Conclusions The predictive model of dissection risk in the acute type A aortic dissection constructed within 72 h after CTA is helpful for the evaluation of the disease and early intervention, and has certain clinical application value. Key words: aortic dissection; rupture; risk factors; risk score; model