Despite the availability of modern imaging technology, 35% of aortic dissections remain undiagnosed in vivo because clinical criteria for aortic dissection are not available to date. The present study analyzed 250 patients with acute chest and/or back pain, absence of an established differential diagnosis of the pain syndrome and clinical suspicion of acute aortic dissection for presence of 26 clinical variables. Multivariate analysis identified an aortic pain syndrome with immediate onset and/or tearing or ripping character (P < 0.0001), mediastinal and/or aortic widening on chest radiography (P < 0.0002) and pulse- and/or blood pressure differentials (P < 0.0001) as predictors of acute aortic dissection. Probability of dissection was low (7%) with absence of all three variables, intermediate (31 and 39%, respectively) with isolated findings of "aortic pain" or "mediastinal widening", and high (> 83%) with either isolated "pulse- and/or blood pressure differentials" or any combination of the three variables. This model appears useful to improve selection of patients for emergency imaging of the thoracic aorta.
