Echocardiographic evaluation should be performed in all patients with clinically suspected IE, including those with negative blood cultures.  Not only is TEE the preferred approach in patients with clinically suspected IE in whom TTE is suboptimal it is also the procedure of choice for imaging the pulmonic valve, patients with PVE (especially at the mitral site) and patients who are at high risk for intracardiac complications or those with sign of persistent or invasive infection despite adequate antimicrobial therapy.

The sensitivity of TTE for the detection of vegetations in NVE is less than 65 per cent, although its specificity is excellent. In contrast, in proven NVE the sensitivity for vegetation detection of TEE was 90 to 100 per cent, and in clinically suspected NVE, it ranged from 82 to 94 per cent. In patients with PVE, TTE is limited by the shadowing effect of the mitral valve prosthesis. The sensitivity of TEE for detecting vegetations in PVE, involving mechanical or bioprosthetic devices ranged from 80 to 96 per cent, whereas that of TTE was from 36 to 16 per cent.   Despite the sensitivity of TEE in detecting vegetations in patients with proven IE, echocardiography does not itself provide a definite diagnosis. Vegetations and valve dysfunction may be demonstrated, but determination of causality requires clinical or direct anatomical and microbiological confirmation.  Infectious vegetation cannot be distinguished from marantic lesions, nor can vegetations be distinguished from thrombus or pannus on prostheses. Further more, it is usually not possible to distinguish active from healed vegetations in NVE. Thickened valves, ruptured chordae or valve calcifications and nodules may be mistaken for vegetations, indicating the specificity limitations of echocardiography.