Temporal Trends in Blood Stream Infection Isolates from Surgical Patients.
BACKGROUND: Blood stream infections (BSIs) are a common source of morbidity and death in hospitalized patients. We hypothesized that the proportions of bacteremia from gram-positive and fungal pathogens have decreased over time, whereas rates of gram-negative bacteremia have increased as a result of better central venous catheter management. METHODS: All U.S. Centers for Disease Control and Prevention-defined BSIs in patients treated on the general surgery and trauma services at our institution between January 1, 1998, and December 31, 2009 were identified prospectively. These cases were analyzed on a yearly basis to compare rates of various infections over time. The Cochran-Armitage test for trend was used to evaluate categorical data, whereas the Jonckheere-Terpstra test for ordered values was used to analyze continuous data. RESULTS: A total of 1,040 patients had 1,441 episodes of BSI caused by 1,632 strains of bacteria or fungi. There was no difference over time in the proportion of BSI among overall infections. Rates of BSI for gram-negative and fungal pathogens increased over time (p=0.03 and<0.0001, respectively), whereas rates of gram-positive BSI decreased (p<0.0001). Positive changes in anaerobic BSI approached statistical significance. CONCLUSION: Although our hypothesis was only partly true, over the last 12 y, our institution clearly has witnessed a shift in the types of organisms causing BSIs. There was a decrease in the rates of BSI caused by gram-positive pathogens with an associated increase in the rates of BSI of infections by fungal and gram-negative pathogens. Interventions to reduce institutional rates of BSI should include targeted therapies based on historical institutional trends.
Guidry, CA; Rosenberger, LH; Petroze, RT; Davies, SW; Hranjec, T; McLeod, MD; Politano, AD; Riccio, LM; Sawyer, RG
Volume / Issue
Start / End Page
Pubmed Central ID
Electronic International Standard Serial Number (EISSN)
Digital Object Identifier (DOI)