Transthoracic echocardiography in assessing patients with suspected endocarditis (TEASE-Forward Study): a prospective cohort study

Background Transesophageal echocardiography (TEE) is a fundamental test for the diagnosis of infective endocarditis (IE). However, there is a high rate of negative TEEs among patients with suspected IE but negative transthoracic echocardiography (TTE), suggesting that TEE might be avoided without clinical consequences in at least a subset of these patients. Purpose In our previous retrospective TEASE study, we developed a risk-stratification model, integrating a negative TTE with clinical and laboratory findings, to identify low-risk patients with suspected IE in whom TEE could be safely deferred or omitted. In this study we aimed to prospectively validate that model in a population of patients referred to undergo TEE for suspected IE but showing negative TTE. Methods Overall, 673 consecutive patients were referred to our echo-lab to undergo TEE for suspected IE. Among these, 383 patients (56.9%) had negative TTE and formed the population of this study. For each patient, we obtained detailed clinical data, including age, sex, blood exams, results of blood cultures (BCs), evidence of systemic/pulmonary embolism, presence of moderate/severe heart valve disease, valve prostheses and cardiac electronic devices (CEDs). Results IE was eventually diagnosed in 63 patients (16.4%). Variables independently associated with IE at multivariate analysis included evidence of embolism (OR 35.49; p<0.001), CEDs (OR 9.90; p<0.001), positive BCs (OR 4.82; p=0.002), and prosthetic heart valves (OR 2.56; p=0.03). Among patients with negative BCs (n=118; 30,8%), IE prevalence increased with the number of other independent predictors of IE, but IE-related mortality was very low even in presence of one or more other independent predictors of IE (1.5%). IE prevalence (4.7%) and IE-related in-hospital mortality (2.4%) were rather low also in patients with positive BCs but absence of any other independent predictors of IE, whereas both IE diagnosis and the related mortality were high in presence of any other independent risk predictor (Table). These data were largely consistent with those found in our previous retrospective study. Conclusion Among patients with clinical suspicion of IE but an initially negative TTE, it is possible to identify low-risk subgroups for both IE and IE-related mortality, in whom TEE may be reasonably deferred or even avoided. Conversely, in high-risk patients, an immediate TEE is warranted to expedite diagnosis and management.