Healthcare-associated vs. hospital-acquired Staphylococcus aureus bacteremia

Jihad Bishara, Elad Goldberg, Leonard Leibovici, Zmira Samra, Hila Shaked, Nariman Mansur, Mical Paul

Research output: Contribution to journalArticlepeer-review

Abstract

Objective: To analyze clinical features and outcomes of patients with hospital-acquired (HA) and healthcare-associated (HCA) Staphylococcus aureus bacteremia. Methods: A retrospective cohort study was conducted from 1988 to 2007. We compared patients with clinically significant HA with those with HCA S. aureus bacteremia. Risk factors for 30-day all-cause mortality were assessed using multivariable logistic regression analysis. Cox regression analysis was used to estimate the hazard ratio (HR) for 5-year mortality with 95% confidence intervals (CI). Results: Of 1261 episodes, 735 (58.3%) were HA and 526 (41.7%) were HCA. The percentage of MRSA was 48.2% (354/735) in HA vs. 42.2% (222/526) in HCA bacteremia; p= 0.04. The percentages of HCA S. aureus bacteremia and MRSA bacteremia did not vary throughout the study period. Mortality at 30 days was 40.2% (507/1261) and at 1 year was 63.4% (800/1261); this was comparable for HA and HCA bacteremia. Five-year survival curves in both settings followed very similar patterns (HR 1.01, 95% CI 0.89-1.15). Risk factors for 30-day mortality were similar, except for primary bacteremia and pre-existing heart valve disease in the HA group. Conclusions: HCA S. aureus bacteremia shares many similarities with HA bacteremia with respect to the prevalence of MRSA strains, mortality rates, and risk factors for death, and should be managed similarly.

Original languageEnglish
Pages (from-to)e457-e463
JournalInternational Journal of Infectious Diseases
Volume16
Issue number6
DOIs
StatePublished - Jun 2012

Keywords

  • Healthcare
  • MRSA
  • Mortality
  • Nosocomial
  • Staphylococcus aureus

Fingerprint

Dive into the research topics of 'Healthcare-associated vs. hospital-acquired Staphylococcus aureus bacteremia'. Together they form a unique fingerprint.

Cite this