Background: The aim of this study was to identify and stratify the most important preoperative factors for in-hospital death after surgery for type A aortic dissection.
Methods: From January 1985 to June 1998, 108 patients underwent surgery for type A aortic dissection. 89.9% of the patients had an acute type A dissection (AD), whereas 11.1% had a chronic dissection (CD). Cardiac tamponade and shock occurred in 22% and 14.8% of the patients, respectively. The location of the primary intimal tear was in the ascending aorta in 71.2% of the cases, in the arch in 16.6% and in the descending aorta in 7.4%. Univariate and multivariate analyses were conducted to identify non-embolic variables independently correlated to in-hospital death. A predictive model of in-hospital mortality was then constructed by means of a mathematical method with the variables selected from logistic regression analysis.
Results: The overall in-hospital mortality rate was 20.3% (22/108 patients), being 9% for CD and 21.6% for AD. Emergent procedures had an in-hospital mortality rate of 47.6%, whereas non-emergent operations had an in-hospital mortality rate of 13.7% (p<0.01). Univariate analysis revealed among 39 preoperative and operative variables, age (years), age >70 years, remote myocardial infarction, cerebrovascular dysfunction, diabetes, preoperative renal failure, shock, cardiopulmonary bypass time (minutes), emergency operation as factors associated to in-hospital death (p<0.05). Stepwise logistic regression analysis selected as independent predicting variables (p<0.05), remote myocardial infarction (p=0.006), preoperative renal failure (p=0.032), shock (p=0.001), age >70 years (p=0.007). Finally, a probability table of death risk was obtained with the logistic regression coefficients. The lower death probability (10.6%) was calculated in absence of risk variables; the higher one in presence of all of them (79.7%). Between these extremes, a total of 64 combinations of death risk were obtained.
Conclusions: Increasing age, shock, coronary artery disease and renal failure are variously associated to a high risk of in-hospital death after surgical correction of type A aortic dissection. This predictive model of death probability allows to collocate preoperatively patients with type A aortic dissection at different levels of risk for in-hospital death.