Relationship between time to positivity of blood culture with clinical characteristics and hospital mortality in patients with Escherichia coli bacteremia.

Background: Previous studies indicated that the time to positivity (TTP) of blood culture is a parameter correlating with degree of the bacteremia and outcome in patients with bloodstream infections caused by Escherichia coli (E. coli). The objective of this study was to further investigate the diagnostic and prognostic power of using TTP to predict E. coli bacteremia.

Methods: A retrospective cohort study at two university hospitals was conducted. We retrieved all the medical records of those with E. coli bloodstream infection according to the records generated by their microbiology departments. Univariate and multivariate analyses were applied to identify clinical factors correlating with fast bacterial growth and significant prognostic factors for hospital mortality.

Results: Medical records of 353 episodes of E. coli bacteremia diagnosed between January 1, 2007 and December 31, 2009 were retrieved in the investigation. Univariate analysis demonstrated that the TTP ≤ 7 hours group is associated with higher incidence of active malignancies (41.7% vs. 27.2%, P = 0.010), neutropenia (30% vs.14.3%, P = 0.007), primary bacteremia (55.0% vs. 33.4%, P = 0.002), and poorer outcome (hospital mortality 43.3% vs.11.9%, P = 0.000) than the TTP > 7 hours group. Multivariate analysis revealed that the significant predictors of hospital mortality, in rank order from high to low, were TTP (for TTP ≤ 7 hours, odds ratio (OR): 4.886; 95% confidence interval (CI): 2.572 - 9.283; P = 0.000), neutropenia (OR: 2.800; 95%CI: 1.428 - 5.490; P = 0.003), comedication of steroids or immunosuppressive agents (OR: 2.670; 95%CI: 0.971 - 7.342; P = 0.057).

Conclusions: Incidence of malignancies, neutropenia and primary bacteremia correlates with fast bacterial growth in patients with E. coli bacteremia. The parameter of TTP has been identified as a variable of highest correlation to hospital mortality and therefore can be potentially utilized as a mortality prognostic marker.