RESUMO
OBJECTIVE: To analyze the peri-operative risk factors of mortality in patients with aortic dissection (AD). METHODS: Between January 2003 and June 2008, 361 AD patients at our hospital were enrolled. Their demographics, history, clinical characteristics and laboratory examinations were retrospectively analyzed. Twenty pre-operative variables were analyzed to identify the predictors of perioperative mortality of AD patients by the analyses of univariate and multivariate logistic regression. RESULTS: The analysis of univariate logistic regression showed that history of hypertension [odds ratio (OR) 0.465, 95% confidence interval (CI) 0.229 - 0.947, P = 0.035], Stanford type A (OR 2.758, 95%CI 1.054 - 7.213 P = 0.039), acute course (OR 7.897, 95%CI 1.874 - 33.275 P = 0.005), neurological symptoms (OR 0.275, 95%CI 0.140 - 0.541, P < 0.001) and operation or not (OR 8.206, 95%CI 4.205 - 16.012, P < 0.001) had a higher mortality in AD patients. The multivariate analysis revealed that acute course (OR 8.178, 95%CI 1.796 - 37.242, P = 0.007), Stanford type A (OR 3.236, 95%CI 1.104-9.487 P = 0.032), neurological symptoms (OR 0.350, 95%CI 0.159 - 0.770, P = 0.009) and operation or not (OR 9.429, 95%CI 4.456 - 19.952, P < 0.001) were significant independent predictors of perioperative mortality in AD patients. CONCLUSION: History of hypertension, acute course, Stanford A and positive neurological symptoms are the independent predictors of perioperative mortality in AD patients. Operation or not is a determinant of patient outcome.