How Much Blood is Too Little for Neonatal Cultures?


Source: Schelonka RL, et al. Volume of blood required to de tect common neonatal pathogens. J Pediatr 1996;129:275-278.

In a recent study, schelonka et al evaluated the minimum volume of blood required to reliably detect low-level bacteremia using the BacT/Alert system, which detects metabolically derived CO2 by a colorimetric method. Recovery of Escherichia coli, Streptococcus agalactiae, Staphylococcus epidermidis, and Candida albicans was evaluated by inoculation of 1-3 colonies into blood. Volumes of this blood, varying from 0.25 mL to 4.0 mL, were then inoculated into 96 blood culture bottles for each organism with subsequent incubation for five days. Analysis of results by the Poison statistic found that a 0.5 mL inoculum was inadequate to reliably detect bacteremia with fewer than 4 colony-forming-units (CFU) per mL of blood.


As the authors point out, previous studies have reported that 23% of neonates with E. coli bacteremia had less than 4 CFU/mL and that low-level bacteremia is encountered frequently in children infected with other organisms as well.1 At the same time, they also point out that it is commonly recommended that 0.5-1.0 mL of blood be obtained for culture in small children and infants. However, their data indicate that a 0.5 mL sample of blood has only an 87% chance of detecting infection in an infant with only 4 CFU/mL of blood. Considering even lower colony counts, they calculate that as many as 60% of blood cultures will yield falsely negative results if only 0.5 mL is sampled. At the same time, they also point out that in very small preterm infants, 0.5 mL accounts for 1.00-1.25% of the total blood volume.

Thus, for infants with an adequate blood volume, a total of at least 1.0 mL of blood must be sampled to have a reliable chance of detecting low-level bacteremia. Better yields will be obtained with larger volumes. (Dr. Deresinski is Clinical Professor of Medicine, Stanford University, and Director, AIDS Program, Santa Clara Valley Medical Center.)


1. Dietzman DE, et al. Neonatal Escherichia coli septicemia: Bacterial counts in blood. J Pediatr 1974; 85:128-130.