Transesophageal echocardiography is an important diagnostic tool available to the critical care physician. Indications for the use of transesophageal echocardiography in the intensive care unite include: critical illness and circulatory shock, thoracic aortic dissection, pulmonary embolism and endocarditis. Probe insertion is easy and is successful in 98% of intensive care patients. Further information concerning 44% of the patients was obtained with transesophageal as compared to transthoracic echocardiography. Transesophageal echocardiography is particularly helpful in evaluating cardiac size and function in patients with circulatory shock. When these patients are on multiple positive inotropic and vasopressor drugs, transesophageal echocardiography is useful in assessing left ventricular preload. In these patients, hemodynamic estimation of left ventricular filling may be misleading. More difficult is the assessment of hemodynamic events in patients requiring mechanical ventilation with increasing positive end-expiratory pressure (PEEP). This ventilation mode develops its own pathophysiology which superposes the effects of the underlying disease. Transesophageal examination of the heart after application of PEEP up to 16 cm H2O demonstrated an acute decrease in size of the right and left ventricle (Figure 1). Cardiac index was affected by the decrease in right ventricular dimension by external compression. The addition of positive and expiratory pressure to a level of 8 cm H2O did not depress cardiac index in patients with severe left ventricular dysfunction [381. PEEP ventilation is associated with abnormal filling patterns characterized by a significant reduction in peak early filling velocity, acceleration and deceleration rate of early filling and peak early to atrial filling velocity ratio (Figures 2a and 2b). Stoddard et al. [41] showed, that these findings of transmitral Doppler flow indices are associated with abnormal left ventricular relaxation. Analysis of regional wall motion under different PEEP levels demonstrated a distinct transmission of increased intrathoracic pressure on the left ventricular wall. We found significant changes in systolic wall motion, in particular a decrease in systolic shortening of the septum and an increase of the lateral wall. Thus, increased intrathoracic pressure under PEEP ventilation is associated with nonuniform regional changes in systolic contraction and abnormal left ventricular relaxation. Both factors are responsible for the decrease in cardiac index under PEEP ventilation. Right ventricular infarction: The transgastric view is usefull in detecting right ventricular wall motion abnormalities and dilatation. Hemodynamically significant right ventricular infarction occurs in the posterior wall, which makes the transesophageal approach ideal. We studied a group of 39 patients with right ventricular infarction. Right infarction was associated in 33 patients with posterior left ventricular infarction (85%) and in three patients with anterior infarction. In two cases only an isolated right ventricular infarction was found. Right ventricular dilatation occurred in 24 patients (61%). Hemodynamic criteria were fullfilled in eleven out of 21 patients (53%). Recognition of regional wall motion abnormalities by transesophageal echocardiography permits an accurate bedside identification of right ventricular infarction (Figure 3). 2D and M-mode transesophageal registration of the short axis improves fight infarction assessment. Pulmonary embolism: Transthoracic echocardiography plays an important role in the diagnostic procedure of patients with pulmonary embolism. When more than 30 to 50% of the cross-section are occluded, pulmonary artery pressure increases and consequently signs of fight heart pressure overload can be visualized by echocardiography: fight ventricular and pulmonary artery dilatation, paradoxical motion of the interventricular septum, left ventricular excentricity. Using the transesophageal approach, the main pulmonary artery, fight and in part, left pulmonary artery can be visualized. In a prospective study Wittlich et al. [43] evaluated the accuracy of transesophageal echocardiography in pulmonary embolism. 109 consecutive patients with suspected pulmonary embolism were evaluated. In 60 patients with signs of fight ventricular overload transesophageal echocardiography was additionally performed. Central pulmonary embolism were found in 35 patients (58%), 29 in the right artery, three in the left and three in both (Figure 4). Pulmonary embolism was the first event in 54% and recurrent in 46%. Sensitivity of transesophageal echocardiography compared to reference methods such as digital substraction angiography, computer tomography, surgery and autopsy was 97%, the specificity 88%, positive predictive accuracy 91% and negative predictive accuracy 96%. According to these findings, transesophageal echocardiography seems to be an accurate and safe tool for detection of central pulmonary embolism. It can avoid further invasive diagnostic procedures in 50% of the patients, making it possible for therapy to be carried out without further delay. Aortic dissection: The transesophageal echocardiography represents the method of choice for the diagnosis of aortic dissection. Echotomographically the whole thoracic aorta, particularly the descending part, can be visualised similarly to computer tomography. Limitations are related to the interposition of the trachea when imaging the aortic arch. The intimal flap separates the false and true lumina. They can be differentiated: 1. The systolic enlargement of the true lumen can be detected by M-mode registration (Figure 5), 2. by the detection of spontaneous echocontrast and thrombus formation in the false lumen, 3. by demonstration of systolic forward flow in the true lumen and delayed or no flow in the false lumen with pulse Doppler echocardiography, 4. by demonstrating entry jets during systole at the entry tear using color Doppler. Thrombus formation is of prognostic value. Thrombosis was observed in type I dissection in 17% and in type III dissection in 40% of the patients and was progressive during follow-up in about 80% of survivors. These patients seem to have a lower mortality. Progressive thrombosis as a sign of healing was observed by autopsy only in 6% of patients with entry tears in the ascending aorta. Patients with free communication and signs of high flow rate in the false lumen have a higher reoperation rate and mortality. The same poor prognosis have patients with tears in the descending aorta and retrograde dissection up to the ascending aorta. Follow-up studies suggest that it is important to detect and resect intimal tears. By a transesophageal approach, intimal tears can be detected in 54% of patients. The sensitivity for transesophageal echocardiography in the diagnosis of aortic dissection was 99% with a specificity of 98%. In contrast, sensitivity and specificity for computer tomography and angiography were 83% and 100%, and 88% and 94%, respectively. Moreover with transesophageal echocardiography, aortic insufficiency and pericardial effusion were detected in 35% and 21% of the patients, respectively. Based on this information, transesophageal echocardiography is the prefered procedure for aortic dissection evaluation and its surgical and medical follow-up. Endocarditis: Reliable identification of endocarditis associated vegetations with transthoracic echocardiography is not satisfactory. Prospective studies evaluated transesophageal echocardiography in patients with suspected endocarditis. Conventional transthoracic echocardiography showed vegetations in 40% of the patients. In contrast, transesophageal echocardiography demonstrated vegetations or diseased valves in 94% of the patients (Figure 6). The studies show a sensitivity of 81.8% and a specificity of 95.2% with positive or negative predictive value of 90% and 90.9%, respectively. When the size of the vegetation was < 5 mm, 6 to 10 mm and > 11 mm, the detection rates with transthoracic echocardiography were 25%, 79%, and 100% respectively as compared with a 100% detection rate for all sizes of vegetations with transesophageal echocardiography. The size of vegetation can be related to clinical signs: 63% of small vegetations were associated with a positive blood culture and 86% to 88% of medium sized or large ones. Transesophageal echocardiography is of value in recognition of subaortic complications in aortic valve endocarditis. In 55 consecutive patients, 24 (44%) had involvement of subaortic structures, including abscesses, perforation into the left atrium or perforation of the anterior mitral leaflet, and intervalvular fibrosa aneurysms. All transesophageal findings were confirmed at surgery and necropsy. Transthoracic examination visualised
