Statistics from Altmetric.com
Editor—In a recent study by Jawaheer et al, the relation between volume of blood cultured and detection rate of coagulase negative staphylococci (CNS) was studied in neonates.1Blood culture bottles inoculated with small blood volumes yielded positive results more frequently than those seeded with larger volumes: 14 of 29 (48%) bottles inoculated with less than 0.5 ml were positive; only seven of 44 (16%) containing ⩾0.5 ml of blood grew the organism (P=0.006). These results are quite unexpected, because large blood volumes should have contained higher number of circulating organisms and, therefore, the positive rate of these bottles should have been higher, but no explanation for this apparent paradox was suggested by the authors.
The overall positivity rate found in the study, and particularly that of the low blood culture bottles, is substantially higher than that found in most blood culture studies, but lack of clinical or laboratory data on the culture positive patients makes it impossible to distinguish contamination of blood cultures from true bacteraemia. This distinction is very important because CNS have not only emerged as the most common pathogens in preterm infants, but are also the major component of the aerobic skin flora of neonates.2
As the authors correctly stated, obtaining large blood culture volumes from small, sick, premature babies is often difficult. Such technical difficulty may result in prolonged manipulation of the skin during attempts to draw the blood sample on one hand, and in failure to obtain an adequate blood volume on the other. I suggest that the high isolation rate of CNS found in blood culture bottles inoculated with small blood volumes probably reflects contamination of the sample with patients’ skin flora.
Editor—In their study of coagulase negative staphylococci in neonatal blood cultures, Jawaheer et al conclude that these organisms can be recovered from blood samples less than 0.5 ml in volume, and that these low volumes do not affect the time taken for cultures to become positive. They described three samples (4.1%) as false negative because a further culture taken within 72 hours was positive.
It is impossible to argue with the first conclusion, but it is also an inescapable fact that small sample volumes increase the probability of false negative results (13% in 0.5 ml samples at 4 CFU/ml).1-1 The only safe conclusion that can be drawn from their observation that the positive cultures were obtained from significantly smaller samples on average than the negative cultures, is that doctors take less blood from sicker infants. It is also likely that more than three of their negative cultures were falsely negative.
We should also expect that larger samples will yield positive results sooner.1-2 Is it really safe to draw the opposite conclusion on the basis of a lack of correlation between volume and time of positive result in 21 samples?
Doctors will continue to take very small blood samples from sick neonates for culture but while this is often unavoidable, no one should imagine that it is “just as good.”
Dr Jawaheer et al reply: We agree with Dr Finn that small blood volumes increase the probability of false negative results. Indeed, as mentioned in our discussion, blood volume is the single most important factor in determining sensitivity of blood cultures. Although some 40% of bottles contained a volume less than that recommended by the manufacturer in our study, the bacterial load in infected infants was sufficiently high to be detected by these small volumes. It may be true that smaller blood volumes were collected from sicker patients. However, all babies in the study had clinical features consistent with infection and no attempt was made to correlate blood volume with severity of clinical symptoms.
False negative results are always difficult to determine. We consider our definition to be reasonably sound: continued clinical evidence of infection with a positive culture drawn 72 hours after the original culture—that is, 24 hours after stopping empirical antibiotic treatment. We accept that other false negatives may have occurred in which the patient responded to 48 hours of empirical treatment, or initially responded and then relapsed at a time beyond our 72 hour cutoff.
Dr Finn states that one would expect larger samples to yield positive results sooner, and Dr Yagupsky that larger samples should contain larger numbers of organisms. We would expect a linear relation between volume of blood cultured and time to detection of positivity only if the density of bacteraemia was uniform in all neonates. This is unlikely to be the case, although quantitative blood cultures were not undertaken in our study.
Dr Yagupsky highlights the point made in our discussion that no attempt was made to distinguish blood culture contaminants from true bacteraemia. This distinction can be important, although our current practice is that all coagulase negative staphylococci isolated from blood cultures of babies with clinical features suggestive of late onset sepsis are considered significant and treated with a minimum five day course of antimicrobial drugs.
Clinicians will continue to inoculate blood culture bottles with very small volumes of blood from sick neonates and this situation is unlikely to change. We re-emphasise our point that paediatric blood culture bottles should be validated for blood volumes less than 0.5 ml.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.