Abstract Background Awareness of the risk factors for infective endocarditis (IE) is essential to early diagnosis and treatment. However, real-world data on bloodstream infections (BSI) and IE epidemiology, their implications for diagnostic screening, and the applicability of typical species concepts by guidelines are scarce. Purpose This study aims to illustrate the link between bloodstream infection (BSI) epidemiology and the risk of IE, leading to better risk stratification and potential optimization of echocardiographic screening. Methods A single-center observational registry between January 2015 and December 2021. All consecutive hospitalized adults with BSIs in the study period were included. The echocardiographic screening for IE and the prevalence of definitive IE diagnosis was assessed. All clinically significant hemoculture with the same microorganism during the same hospitalization was considered as one BSI. Results From 4939 BSIs related to 4242 admissions in 3708 patients analyzed 74 (64-83) years; 2882 (58.7%) male, 2.7% had a diagnosis of definite IE. The most prevalent BSI pathogens were Enterobacterales (48.3%) and Staphylococcus aureus (11.5%) (Figure 1). Some less frequent BSI presented the highest risks for IE: Streptococcus bovis complex (IE prevalence: 19.2%), Listeria monocytogenes (16.7%), and Enterococcus faecalis (15.5%). BSI in patients with intracardiac prosthetic material (ICP) showed an IE prevalence almost five times higher than overall BSI. The IE risk was underestimated when microorganisms were aggregated according to the main guidelines (Figure 1). Echocardiography screening (Figure 2) was highest for Staphylococcus aureus (75%), Staphylococcus lugdunensis (72.2%), and Streptococcus bovis complex (69.9%). Although the IE prevalence was high in the Listeria monocytogenes, Enterococcus faecalis, and HACEK BSIs, echocardiography screening was performed from 41.7% to 66.3%. The echocardiography screening was higher in the ICP presence, ranging from 40 to 100% of BSI with this condition. Among BSI without IE cases, it was performed in 27.3% to 59.4%. In a cox multivariate analysis adjusted for gender, age of more than 75 years: other streptococci, HACEK, Streptococcus bovis complex, E. faecalis, oral streptococci, and S. aureus were at significantly higher risk of IE than Other Enterococci. In this adjusted model, the presence of a prosthetic valve (OR 6.73 95%IC 4.49-10.07) and intracardiac device (2.04 1.26-13.25) remained significant risk factors for IE. Conclusions The study illustrates a discrepancy between the prevalence of BSI pathogens and their IE risk and underlines the importance of sufficient granularity in defining species typically linked to IE. Typical microorganism concepts can underestimate high risk when aggregating species or without considering the IE risk factor context. Considering these new insights, echocardiography screening in BSI to detect IE could be optimized.
Oliveira et al. (Sat,) studied this question.