Development and Validation of a Risk Prediction Score For Enterococcal Endocarditis. (4th October 2017)

• Record Type: Journal Article
• Authors: Mattar, Caline; Thomas, Benjamin S; Strasserking, Fiona; Bakeer, Faris; Chung, Jeffrey; Warren, David
• Is Part Of: Open forum infectious diseases. Volume 4(2017)Supplement 1
• Page Start: S549
• Page End: S550
• Publication Date: 2017-10-04
• Subjects: Communicable diseases -- Periodicals; Medical microbiology -- Periodicals; Infection -- Periodicals; 616.9
• DOI: 10.1093/ofid/ofx163.1428
• Languages: English
• ISSNs: 2328-8957

Abstract

Abstract: Background: Enterococci are the third most common cause of infectious endocarditis (IE). The NOVA score was developed in Spain to identify patients at low risk for enterococcal IE for whom transesophageal echocardiography is unnecessary. The score is based on the number of positive cultures, source of infection, prior history of valve disease and presence of a murmur. It has not been validated in other patient populations. Methods: We conducted a retrospective cohort study of patients with enterococcal bloodstream infection (EBSI) during 2008–2012 in a large tertiary hospital in the United States. Our primary outcome was IE, defined using the modified Duke's criteria. Results: Our cohort consisted of 1117 patients with EBSI. The median age was 59 years (range 18–99); 590 (52.8%) were male; 129 (11.5%) were stem cell transplant patients, and 154 (13.8%) had valvular disease. The source of EBSI was: genitourinary in 91 subjects (8.1%), abdominal in 428 (37.8%), and unknown in 604 (54.1%). 581(52%) of ESBI patients had transthoracic echocardiography, 144(12.9%) had transesophageal, and 114(19.6%) had both. 117 (9.5%) of EBSI patients had endocarditis. Among individual components of the NOVA score, the independent IE risk factors were: prior valvular disease [odds ratio (OR) 1.01; 95% confidence interval (CI) 0.56–1.82], presence of a heart murmur (OR 11.8; CI 6.53–18.78), unknown origin of bacteremia (OR 4.02; CI 2.35–6.86), and number of positive blood cultures (OR 2.18; CI 1.35–3.58). The median NOVA score was 4 (interquartile range 0–5) (Figure 1). Among patients with a NOVA score <4, 40 (5.1%) had IE vs. 77 (23.7%) in with NOVA >4, ( P < 0.001). The NOVA score explained 30.5% of the variability in the observed outcome of endocarditis. The Receiver-Operator Characteristic Area Under the Curve (AUC) for the NOVA score was 0.77 (95% CI 0.72–0.81; P < 0.001). Among 649 monomicrobial EBSI patients, 73 (11.2%) had IE; the ROC in this group was 0.76 (CI 0.69–084; P < 0.001). Conclusion: 9.5% of patient in our EBSI cohort had endocarditis. In our study population, 5.1% patients with a NOVA score <4 had endocarditis. The ROC for NOVA score was similar for all EBSI patients and those with monomicrobial infections. Disclosures: All authors: No reported disclosures.