Risk factors for persistent enterococcal bacteraemia: a multicentre retrospective study

Abstract

Conditions favouring persistent enterococcal bacteraemia (p-EB) have not been fully investigated yet. The aim of our study is to analyse risk factors for p-EB and its impact on mortality. International two-centre retrospective study of all hospitalised adults with enterococcal bacteraemia managed with follow-up blood cultures (BCs) during the period 2011-2019. Exclusion criteria were: (1) death within 72 hours from index BCs and (2) polymicrobial bacteraemia. Primary endpoint was p-EB, defined as further isolation of the same species of Enterococcus spp. from BCs after at least 72 hours of appropriate antibiotic therapy. Multivariable logistic regression model was performed to assess risk factors for p-EB. The impact of p-EB on 30-day mortality was assessed by Kaplan-Meier survival curve and Cox regression multivariable model. During the study period, 244 enterococcal bacteraemia were diagnosed. P-EB were 13.5% (33/244). At multivariable analysis, factors independently associated with p-EB were hematologic malignancy (OR 4.60 [95% CI 1.32-16.00], P=0.01), infective endocarditis (OR 7.99 [95% CI 2.20-28.9], P=0.002), and use of daptomycin as initial treatment (OR 4.50 [95% CI 1.29-15.61], P=0.018). Mortality rate was higher in the p-EB group (32% vs. 18%). Kaplan-Meier survival curve showed that patients with p-EB were less likely to survive at 30 days from index BCs (log-rank P=0.002). Using a Cox regression model, independent predictors of 30-day mortality were hematologic malignancy (HR 2.30 [95% CI 1.02-4.11], P=0.043), p-EB (HR 1.93 [95% CI 0.92-4.04], P=0.08), and septic shock (HR 5.92 [95% CI 2.17-16.30], P=0.001). P-EB was diagnosed mainly in very fragile patients and in those receiving daptomycin as frontline therapy. P-EB may have an impact on mortality.