Predictors of nosocomial bloodstream infections in older adults

Keith S. Kaye, Dror Marchaim, Ting Yi Chen, Teena Chopra, Deverick J. Anderson, Yong Choi, Richard Sloane, Kenneth E. Schmader

Research output: Contribution to journalArticlepeer-review

46 Scopus citations


OBJECTIVES: To identify predictors and construct a prediction model for nosocomial bloodstream infection (BSI) in older adults. DESIGN: Retrospective case-control study. SETTING: Hospitals belonging to the Duke Infection Control Outreach Network. PARTICIPANTS: Patients age 65 and older with a nosocomial BSI and matched uninfected controls. MEASUREMENT: Multiple variables were captured and compared between groups. Independent predictors were identified using conditional logistic regression. A prediction model and score was constructed. RESULTS: Eight hundred thirty cases were compared with 830 controls. Eighty-one percent of nosocomial BSIs were catheter related (CRBSI). Methicillin-resistant Staphylococcus aureus (MRSA) was the most common pathogen isolated (23%). Independent predictors of nosocomial BSI in older adults were male sex, obesity, low McCabe score on admission, presence of a central line at admission, gastrostomy at admission, recent surgery, and urinary incontinence. A prediction model score greater than 11 (total possible score 23) was predictive of infection. CONCLUSION: MRSA is a common cause of CRBSI in older adults. Male sex, obesity, the presence of a central line, a gastrostomy tube, and urinary incontinence at the time of admission were independent predictors of BSI in hospitalized older adults. The prediction model constructed in this study should be validated prospectively in a different cohort.

Original languageEnglish (US)
Pages (from-to)622-627
Number of pages6
JournalJournal of the American Geriatrics Society
Issue number4
StatePublished - Apr 2011


  • BSI
  • MRSA
  • bacteremia
  • elderly
  • nosocomial
  • predictors
  • risk factors

ASJC Scopus subject areas

  • Geriatrics and Gerontology


Dive into the research topics of 'Predictors of nosocomial bloodstream infections in older adults'. Together they form a unique fingerprint.

Cite this