Epidemiology, microbiology and outcomes of healthcare-associated and community-acquired bacteremia: A multicenter cohort study

Document Type


Publication Title

Journal of Infection


Objectives: Classically, infections have been considered either nosocomial or community-acquired. Healthcare-associated infection represents a new classification intended to capture patients who have infection onset outside the hospital, but who, nonetheless, have interactions with the healthcare system. Regarding bloodstream infection (BSI), little data exist differentiating healthcare-associated bacteremia (HCAB) from community-acquired bacteremia (CAB). We studied the epidemiology and outcomes associated with HCAB. Methods: We conducted a multicenter, retrospective chart review at 7 US hospitals, of consecutive patients admitted with a BSI during 2006, who met pre-defined selection criteria. We defined HCAB as a BSI in a patient who met ≥1 of the criteria: 1) hospitalization within 6 months; 2) immunosuppression; 3) chronic hemodialysis; or 4) nursing home residence. The rest were classified as CAB. We examined patient demographics, severity of illness, and in-hospital mortality rates by HCAB vs. CAB status. A bootstrap logistic regression model was developed to quantify the independent association between HCAB and hospital mortality. Results: Of the total 1143 patients included, HCAB accounted for 63.7%, with the percentage ranging from 49.0% to 78.1% across centers. HCAB patients were older (58.5 ± 17.5 vs. 55.0 ± 19.9 years, p = 0.003) and slightly more likely to be male (56.1% vs. 50.2%, p = 0.044) than those with CAB. HCAB was associated with a higher mean Acute Physiology Score (12.6 ± 6.2 vs. 11.4 ± 5.7, p = 0.009) and recent hospitalization was the most prevalent criteria for defining HCAB (76.5%). Hospital LOS was longer in the HCAB (median 8, IQR 5-15 days) than CAB (median 7, IQR 4-13 days) group (p = 0.030). In a multivariable model, the risk of hospital death was 3-fold higher for HCAB compared to CAB (adjusted odds ratio 3.13, 95% CI 1.75-5.50, p < 0.001, AUROC = 0.812). Conclusions: HCAB accounts for a substantial proportion of all patients with BSIs admitted to the hospital. HCAB is associated with a higher mortality rate than CAB. Physicians should recognize that HCAB is responsible for many BSIs presenting to the hospital and may represent a distinct clinical group from CAB. © 2010 The British Infection Association.

First Page


Last Page




Publication Date