Inaccuracy of transthoracic echocardiography for the identification of right-sided vegetation in patients with no history of intravenous drug abuse or cardiac device insertion.

OBJECTIVE: The use of transthoracic echocardiography (TTE) to identify right-sided infective endocarditis (RSIE) vegetation is controversial. Data are scarce for patients with no history of intravenous drug abuse (IVDA) or cardiac device insertion. This study analysed the consistency of presurgical echocardiographic results with surgical findings for vegetation identification, and the factors that influence accuracy of echocardiography.

METHODS: This retrospective trial divided infective endocarditis (IE) patients into three subgroups according to the results of their presurgical TTE: left-sided native IE (LSNIE), left-sided prosthetic valve IE (LSPIE) and RSIE. The accuracy of TTE was tested by comparing vegetation (number and location), detected presurgery by TTE, with actual findings during surgery.

RESULTS: In total, 416 patients were analysed, 322 with LSNIE, 31 with LSPIE and 63 with RSIE. Consistency between TTE findings and surgical results was lower in the RSIE group compared with the LSPIE and LSNIE groups. Consistency was lowered by the presence of vegetation in multiple locations and atypical distribution--both of which were increased in the RSIE group. The chance of vegetation in both sides of the heart rose with increased numbers of vegetation locations in RSIE patients. A high proportion of RSIE patients had congenital heart defects, mostly ventricular septal defects.

CONCLUSIONS: TTE may be unsuitable for RSIE patients with no history of IVDA or cardiac device insertion, because multifocal and atypically distributed vegetation may influence detection accuracy.