First, the reanalysis of our recent article by Dr Ghosh is greatly appreciated, because both the importance and the renaissance of clinical medicine, with an emphasis on precision and accuracy, are well documented. Dr Ghosh has converted the data presented in Table 1 of our article into positive and negative LRs, leading him to conclude that in patients with intermediate and high probability (risk) of dissection, the absence of classic clinical signs (such as tearing/ripping pain, mediastinal aortic pain, and mediastinal/aortic widening) would not sufficiently rule out the diagnosis. Moreover, he deduces that even in high- and intermediate-risk patients, the clinical triad of aortic pain, mediastinal/aortic widening, and new-onset aortic regurgitation is not sufficient to justify no additional imaging. We believe that Dr Ghosh was obviously intrigued by the presented results and their simplicity. In addition, the data show another aspect: that the clinical algorithm describes the relative likelihood of a person at risk and encourages confirmatory (diagnostic) imaging tests at any level of probability (even in the low-risk group), less urgently, however. The data are not to be used in the intention to avoid diagnostic imaging as the result of the suggested level of risk, but rather as an emergency triage prior to confirmatory imaging; suggested diagnostic modalities are spiral computed tomography, magnetic resonance imaging, or transesophageal echocardiography, rather than angiography, because of both the inherent risk and the fact that luminograms are more likely to miss difficult diagnoses, such as intramural hematoma, an important precursor of dissection.
Nienaber CA. Clinical Diagnosis of Acute Aortic Dissection—Reply. Arch Intern Med. 2001;161(20):2507. doi: