Clinical predictors and clinical prediction rules to estimate initial patient risk for infective endocarditis in Staphylococcus aureus bacteraemia: a systematic review and meta-analysis

Abstract

Objectives: We conducted a meta-analysis to summarize diagnostic properties of risk factors and clinical prediction rules for diagnosing infective endocarditis (IE) in Staphylococcus aureus bacteraemia (SAB).

Methods: We searched MEDLINE, Embase, and the Cochrane Database from inception to 6 January 2016 to identify studies evaluating risk factors and clinical prediction rules for IE in SAB patients. Pooled estimates of diagnostic properties for main risk factors were calculated using a bivariate random effects model.

Results: Of 962 articles identified, 30 studies were included. These involved 16 538 SAB patients including 1572 IE cases. Risk factors with positive likelihood ratio (PLR) greater than 5 included embolic events (PLR 12.7, 95% CI 9.2-17.7), pacemakers (PLR 9.7, 95% CI 3.7-21.2), history of previous IE (PLR 8.2, 95% CI 3.1-22.0), prosthetic valves (PLR 5.7, 95% CI 3.2-9.5), and intravenous drug use (PLR 5.2, 95% CI 3.8-6.9). The only clinical factor with negative likelihood ratio (NLR) less than 0.5 was documented clearance of bacteraemia within 72 hours (NLR range 0.32-0.35). Of the nine published clinical prediction rules for ruling out IE, five had an NLR below 0.1.

Conclusions: SAB patients with high-risk features (embolic events, pacemakers, prosthetic valves, previous IE, or intravenous drug use) should undergo a trans-esophageal echocardiography (TEE) for IE. Clinical prediction rules show promise in safely ruling out endocarditis, but require validation in future studies.

Keywords: Clinical prediction rules; Infective endocarditis; Meta-analysis; Sensitivity and specificity; Staphylococcus aureus bacteraemia.