Predictors of Vancomycin-Resistant Enterococcus (VRE) Bacteremia in Ontario, Canada

Abstract

BackgroundTo determine predictors of vancomycin resistant enterococcus (VRE) bacteremia in Ontario, Canada.MethodsOntario hospitals are mandated to report VRE bacteremias to a public reporting database. All confirmed VRE bacteremias between January 2009 - December 2013 were linked to provincial health care administrative data sources. A population-based, nested case–control study was performed to determine predictors of VRE bacteremia. Cases were patients with VRE bacteremia and controls were patients with at least one hospital admission during the study period. Each case was matched with up to three controls using frequency matching on age, sex and aggregated diagnosis group. Associations between patient- and hospital-level predictors and VRE bacteremia were estimated by Generalized Estimating Equations and summarized using odds ratios (OR) (adjusted for age, sex, Charlson score, intensive care unit (ICU) admission, length of stay, hospital admission, comorbidities, hospital size and hospital type) and corresponding 95% confidence intervals (CI) in SAS.ResultsIn total, 232 patients had a VRE bacteremia during the study period; 217 cases were successfully linked to administrative data sources and there were 651 controls. Mean age of cases was 61 years (SD 17) vs. 60 years for controls (SD 21). The proportion of male cases and controls was 60%. Length of stay for cases was longer than controls (median 39 days [range 1–539 days] vs. 3 days [range 1 – 136 days], P < 0.001) and 82% of cases died within 30 days vs. 21% of controls (P < 0.001). In adjusted analyses, patient-level predictors of VRE bacteremia included: organ transplant (OR 18.93 [95% CI 8.37 – 42.79), cancer (OR 9.56 [95% CI 4.61 – 19.79]), ICU admission (OR 7.45 [95% CI 3.57 – 15.54]), heart disease (OR 5.03 [95% CI 1.92 – 13.18]) and length of stay (OR 1.08 per day [95% CI 1.03 – 1.12]); COPD (OR 3.10 [95% CI 0.86 – 11.20]) and diabetes (OR 2.35 [95% CI 0.72 – 7.64]) were not significant predictors. Hospital-level predictors included hospital size ≥800 beds (OR 10.64 [95% CI 2.34 – 48.25]) and teaching hospitals (OR 4.20 [95% CI 1.65 – 10.74]).ConclusionImmunocompromised and patients admitted to ICU are at highest risk of VRE bacteremia, particularly at large hospitals and teaching hospitals. These results may help inform clinical decisions and infection prevention programs.Disclosures All authors: No reported disclosures.