Estimating bacterial load in S. aureus and E. coli bacteremia using bacterial growth graph from the continuous monitoring blood culture system.

Abstract

We examined whether the time to positivity (TTP) and growth and detection plot graph (GDPG) created by the automated blood culture system can be used to determine the bacterial load in bacteremic patients and its potential association correlation with disease severity. Known bacterial inocula were injected into the blood culture bottles. The GDPGs for the specific inocula were downloaded and plotted. A cohort of 30 consecutive clinical cultures positive for S. aureus and E. coli was identified. Bacterial load was determined by comparing the GDPG with the "standard" curves. Variables associated with disease severity were compared across 3 bacterial load categories (< 100, 100-1000, > 1000 CFU/mL). S. aureus growth was sensitive to the blood volume obtained whereas E. coli growth was less so. A 12-hour delay in sample transfer to the microbiology laboratory resulted in a decrease in TTP by 2-3h. Mean TTP was 15 and 10h for S. aureus and E. coli, respectively, which correlates with > 1000 CFU/mL and 500-1000 CFU/ml. For S. aureus, patients with a bacterial load > 100 CFU/mL had a higher mortality rate, (OR for death = 9.7, 95% CI 1.6-59, p = 0.01). Bacterial load > 1000 CFU/mL had an odds ratio of 6.4 (95% CI1.2-35, p = 0.03) to predict an endovascular source. For E. coli bacteremia, we did not find any correlations with disease severity. GDPG retrieved from the automated blood culture system can be used to estimate bacterial load. S.aureus bacterial load, but not E.coli, was associated with clinical outcome.