Transthoracic echocardiography versus transesophageal echocardiography in detecting cardiac sources of embolism.

Although the yield of potential cardiac sources of embolism by echocardiography in patients with stroke and arterial embolism has been low, with the advent of transesophageal echocardiography, a renewed enthusiasm for echocardiography in these patients has developed. This article reviews the six major studies comparing transthoracic to transesophageal echocardiography in the search for potential cardiac sources of embolism. The overall yield of transesophageal echocardiography in these studies for potential cardiac sources of embolism is 43% compared to 14% by transthoracic echocardiography in a total of 367 patients. In patients without clinical cardiac disease, the yield is lower but still substantially higher by transesophageal echocardiography (24% compared to 7% by transthoracic echocardiography). For left atrial thrombus, left atrial spontaneous contrast, patent foreman ovale, and atrial septal aneurysm (ASA), transesophageal echocardiography is clearly superior than transthoracic echocardiography. Data on the detection of mitral valve prolapse and left ventricular thrombus are conflicting and neither method is clearly superior. In addition, transesophageal echocardiography identifies certain abnormalities including debris in the aorta and prosthetic strands that transthoracic echocardiography is incapable of identifying. Although transthoracic echocardiography should continue to be the initial screening modality for stroke patients, transesophageal echocardiography should be performed when surface findings are negative or equivocal in patients with likely cardioembolic stroke.