(P166) INCIDENCE AND PREDICTORS OF INFECTIVE ENDOCARDITIS IN BACTEREMIC PATIENTS

Background: Infective endocarditis (IE) is a serious disease, with a high mortality rate. Diagnosis of IE is made either pathologically or with a combination of clinical, microbiological, and imaging criteria. There has been a shift in the composition of organisms causing infective endocarditis. We aimed to provide a contemporary incidence of IE among bacteremic patients stratified by bacterial pathogens. We also sought to evaluate the predictors of IE in bacteremic patients.

METHODS AND RESULTS: We included patients at a large quaternary centre in Canada. All bacteremias during the period from 2014 to 2016 were identified and data was collected on the following variables – patient demographics, organism, comorbidities, pre-disposing factors, clinical findings, pertinent biochemistry, and diagnosis of IE. Diagnosis of IE was made using the modified Duke’s criteria. Logistic regression was used to identify predictors of IE.
There were 1,708 patients with bacteremia included in this analysis. The mean age was 64.5±17.0 years old; 39.7% were females and 77.4% were community acquired bacteremia. In addition, 47.0% had hypertension, 29.9% had diabetes mellitus, 17.5% had chronic kidney disease and 25.2% were immunosuppressed.
There were 102 (6.0%) patients with confirmed infective endocarditis (IE). The proportion of bacteremic patients who developed IE was 13.2% in 281 patients with methicillin-sensitive Staphylococcus aureus, 5.4% in 56 patients with methicillin-resistant S. aureus, 2.4% in 292 patients with S. epidermidis, 9.9% in 121 patients with Streptococcus viridans group, 16.5% in 91 patients with Enterococcus, 14.3% in 7 patients with HACEK organisms, 2% in 47 patients with fungal bacteremia, 1.3% in 303 patients with Escherichia coli, 38.5% in 13 in patients with Streptococcus bovis, 9.5% in 21 patients with Propionibacterium acnes, 3.0% in 33 patients with Streptococcus pneumoniae, and 1.9% in 108 patients with Klebsiella bacteremia.
Predictors of IE were intravenous drug use (odds ratio [OR]=8.64, 95% confidence interval [CI] 3.50- 21.3; p< 0.0001), prosthetic heart valve (OR= 2.10, 95% CI 1.01- 4.39; p=0.049), structural heart disease (OR= 12.2, 95% CI 3.35- 44.67; p= 0.0002) and community acquired bacteremia (OR= 2.06, 95% CI 1.11- 3.82; p=0.02).

Conclusion: Methicillin-sensitive Staphylococcus aureus is the commonest cause overall of IE followed by viridans group streptococcus. Streptococcus bovis and Enterococcus bacteremia have the highest incidence of IE. Knowledge of the incidence rates of IE among various pathogens may guide diagnostic testing such as pursuing more advanced imaging procedures for patients with increased risk of IE. Further studies are needed to identify the optimal screening strategy for IE.