Diagnostic accuracy and therapeutic impact of transthoracic and transesophageal echocardiography in mechanically ventilated patients in the ICU

Abstract

Study objectives: To assess the respective diagnostic accuracy of transthoracic echocardiography (TTE) and transesophageal echocardiography (TEE) and their therapeutic implications in mechanically ventilated patients, in the intensive care unit (ICU).

Design: A prospective study.

Settings: Intensive care units of two tertiary referral teaching hospitals.

Patients: One hundred eleven ICU patients (81 men and 30 women; mean age 57 +/- 16 years). Fifty-seven percent were hospitalized for medical illnesses, 16.5 percent after thoracic surgery, 10.5 percent after other surgery, and 16.0 percent for multiple trauma. Their Simplified Acute Physiologic Score was 16 +/- 5.

Interventions: The echocardiograms were performed in order to solve well-defined clinical problems. TTE was the first step of the procedure and TEE was performed only when (1) TTE did not solve the clinical problems, and (2) TTE yielded unsuspected findings requiring TEE. During each echocardiographic study, the following were noted: ventilatory mode, clinical problems, imaging quality, results, consequence on acute care, duration of the procedure, and potential complications of TEE. Diagnostic accuracy was defined as the proportion of solved problems, and therapeutic impact was defined as changes on acute care that resulted directly from the procedure.

Measurements and results: One hundred twenty-eight consecutive TTE and 96 TEE were performed. TTE solved 60 of 158 clinical problems (38 percent), whether positive end-expiratory pressure (> 4 cm H2O) was present or not (28 of 74 vs 32 of 84: p > 0.50). TTE allowed evaluation of left ventricular function in 77 percent of cases and pericardial effusion in every case, but it did not solve most of the other clinical problems. Indeed, the diagnostic accuracy of TEE was markedly superior (95/98 vs 60/158: p < 0.001), but TEE required a physician's presence longer (43 +/- 17 min vs 27 +/- 12 min: p < 0.001). When TTE and TEE were scheduled (n = 96), TEE yielded an additional diagnosis or excluded with more certitude a suspected diagnosis, except in two cases. TEE had a therapeutic impact more frequently than TTE (35/96 vs 20/128: p < 0.001). Cardiovascular surgery was prompted by echocardiographic findings in ten patients. TEE was well tolerated in all patients; there were no complications.

Conclusions: TEE is a valuable well-tolerated imaging technique in mechanically ventilated patients. For the assessment of left ventricular systolic function and pericardial effusion; however, TTE continues to be an excellent diagnostic tool, even when positive end-expiratory pressure is present. Both TTE and TEE have a therapeutic impact in approximately 25 percent of cases.