Context The diagnosis of acute thoracic aortic dissection is difficult to make and often missed.
Objective To review the accuracy of clinical history taking, physical examination, and plain chest radiograph in the diagnosis of acute thoracic aortic dissection.
Data Sources A comprehensive review of the English-language literature was conducted using MEDLINE for the years 1966 through 2000. Additional sources were identified from the references of retrieved articles.
Study Selection The search revealed 274 potential sources, which were reviewed for pertinence and quality. Articles included were original investigations describing the clinical findings for 18 or more consecutive patients with confirmed thoracic aortic dissection. Twenty-one studies were identified that met selection criteria.
Data Extraction Critical appraisal and data extraction were performed by the author.
Data Synthesis Most patients with thoracic aortic dissection have severe pain (pooled sensitivity, 90%) of sudden onset (sensitivity, 84%). The absence of sudden pain onset lowers the likelihood of dissection (negative likelihood ratio [LR], 0.3; 95% confidence interval [CI], 0.2-0.5). On examination, 49% of patients have an elevated blood pressure, 28% have a diastolic murmur, 31% have pulse deficits or blood pressure differentials, and 17% have focal neurological deficits. Presence of a diastolic murmur does little to change the pretest probability of dissection (positive LR, 1.4; 95% CI, 1.0-2.0), whereas pulse or blood pressure differentials and neurological deficits increase the likelihood of disease (positive LRs, 5.7 and 6.6-33.0, respectively). The plain chest radiograph results are usually abnormal (sensitivity, 90%); hence, the presence of a normal aorta and mediastinum decreases the probability of dissection (negative LR, 0.3; 95% CI, 0.2-0.4). Combinations of findings increase the likelihood of disease.
Conclusions The presence of pulse deficits or focal neurological deficits increases the likelihood of an acute thoracic aortic dissection in the appropriate clinical setting. Conversely, a completely normal chest radiograph result or the absence of pain of sudden onset lowers the likelihood. Overall, however, the clinical examination is insufficiently sensitive to rule out aortic dissection given the high morbidity of missed diagnosis.
