The MEFIER Score-A Risk Score to Stratify Infective Endocarditis in Patients With Bacteremia Based on an 11-Year Territory-Wide Cohort.

BACKGROUND

The 2023 Duke-International Society for Cardiovascular Infectious Diseases (ISCVID) guideline included as a typical infective endocarditis (IE) organism irrespective of the primary source and setting of infection. Universal echocardiogram for bacteremia is resource-intensive. Our aim was to develop a risk score that can be applied at the time of bacteremia.

METHODS

We conducted a retrospective cohort study utilizing a territory-wide electronic database and included all inpatients (age ≥18) between 2009 and 2019 who had at least 1 blood culture positive for . A multistep machine learning algorithm was employed to generate the risk score.

RESULTS

A total of 2535 distinct bacteremia episodes were analyzed. The percentage of IE was 3.39% (86/2535). The Machine Learning-derived Infective Endocarditis Risk (MEFIER) score was developed based on history of valvular heart disease (28 points), history of congenital heart disease (14 points), presence of cardiac implantable electronic device (12 points), age (≥18 to <43 years: 3 points; ≥43 to <65 years: 6 points), male gender (2 points), community onset (11 points), and abnormal levels of hemoglobin (16 points), albumin (3 points), and platelets (1 point). With a threshold of ≥32 indicating high risk for IE, the MEFIER score demonstrated an area under the receiver operating characteristic curve of 0.79 (95% CI, 0.72-0.88) and a negative predictive value of 0.98 (95% CI, 0.97-0.99).

CONCLUSIONS

The MEFIER score is a unique tool to stratify IE patients with bacteremia, boasting a negative predictive value of 98%, to guide use of echocardiography in these patients.