Background and purpose

Blood culture is important for the diagnosis of sepsis, but it is sometimes difficult to differentiate true bacteremia from pseudobacteremia. This study proposed clinical criteria and evaluated the role of repeat blood cultures in assessing the clinical significance of potential contaminants in blood cultures (PCBCs).

Methods

From February to May in 2004 (prospective study) and 2003 (retrospective study), adult patients with growth of coagulase-negative staphylococci, Bacillus spp., Micrococcus spp., Propionibacterium spp., Gram-positive bacilli, or Clostridium perfringens, collectively referred to as" PCBCs", in at least 1 set of blood cultures in a medical center were included. The demographic and clinical data of patients with PCBCs were collected, and proposed clinical criteria for true bloodstream infections were used to evaluate their clinical outcome. Also, the potential role of repeating blood cultures to differentiate true bacteremia from pseudobacteremia was evaluated.

Results

There were 212 cases with 214 PCBCs, of which coagulase-negative staphylococci predominated (182 isolates, 85.0%). The overall contamination rate was 83.9%(178/212). Repeating 2 sets of blood cultures might be useful in the clinical differentiation of true bacteremia and pseudobacteremia, since the contamination rate of patients with potential contaminants in 1 set of blood cultures declined from 95% to 87%(p= 0.04) with such a strategy. Those with true bloodstream infections had a significantly higher all-cause mortality rate at 14 days than those with pseudobacteremia (23.8% vs 7.3%, p= 0.028), suggesting the validity of the clinical criteria. Of the 178 cases with pseudobacteremia, 73 (41.0%) were unnecessarily treated by systemic antibiotics, of which glycopeptides accounted for 20.0%. For these cases, antimicrobial therapy offered no survival benefit.

Conclusions

In an era of increasing glycopeptide resistance among Gram-positive cocci, clinical strategies for the early diagnosis of pseudobacteremia in cases with PCBCs are urgently required, in order to avoid the unnecessary use of glycopeptides. The proposed criteria and repeat blood culturing seem to be useful in the assessment of the clinical significance of PCBCs, and for reduction of the inappropriate use of glycopeptides.