A Normal Transthoracic Echocardiogram can be Used to Rule Out Infective Endocarditis in Patients with Staphylococcus aureus Bacteremia
Methods: We retrospectively evaluated hospitalized patients with S. aureus bacteremia (SAB) from seven hospitals in Toronto, Ontario over a 3-year period. In 536 patients who received a TTE within 28 days of bacteremia, we randomly divided the sample into development (n=268) and validation (n=268) cohorts. We derived criteria to rule out infective endocarditis based on multivariable logistic regression analysis in the development cohort, and applied these criteria to the validation cohort to determine criteria diagnostic properties.
Results: In multivariable analysis, four criteria were useful to rule out infective endocarditis: a normal TTE (p<0.001), non-community-acquired bacteremia (p=0.002), no intravenous drug use (p<0.001), and absence of high-risk cardiac conditions (p<0.001). In the validation sample, failing any criteria had a 97% sensitivity, 52% specificity, 25% positive predictive value, and 99% negative predictive value for infective endocarditis. The negative likelihood ratio was 0.05.
Conclusion: A prediction model including normal TTE, non-community acquired bacteremia, no intravenous drug use, and absence of high-risk cardiac conditions is a useful tool to identify patients at low risk of infective endocarditis. Risk stratification combined with TTE may be adequate to rule out IE in most patients with SAB.
A. Bai, None
D. Ricciuto, None
M. Steinberg, None
T. Fernandes, None
A. Chiu, None
S. Raybardhan, None
E. Ferndando, None
C. Bell, None
A. Morris, None
See more of: Poster Abstract Session