Introduction Blood culture results guide appropriate treatment of infections, choice of antibiotic and length of course, all essential in antibiotic stewardship. Results with high sensitivity and specificity are vital, and contaminated samples pose challenges in clinical practice.

Aim To determine incidence of blood culture positivity and contamination in our hospital, and compare time to positivity for growths deemed significant or contaminants.

Methods We undertook a retrospective study of all blood cultures taken from children attending our children’s hopsital over two 3 month periods in 2014 and 2015. Samples from intensive care and oncology were excluded due to high rates of central venous access devices. BacT-Alert incubation and MAKDI-Tof systems were used to identify organisms. Positive cultures were classed as significant or contaminants depending on microbiology opinion and clinical management of the patient.

Results 1270 blood cultures were included, of which 92 (7.2%) were positive; 34 (37%) of which were deemed significant (2.7% of total).

Conclusions While the overall rate of positivity is similar to published data¹, the proportion of contaminated samples is higher than optimal (3%²). The median age of patients with contaminants isolated was significantly lower than those with significant growths. A contributing factor might be that taking samples from unwell infants can be difficult and samples may therefore be small. It is well-recognised that the yield of blood cultures is highly dependent on blood volume, and contamination rates are higher in smaller samples^1,2. We aim to highlight the importance of optimising blood volume and technique, to reduce contamination rates. Decisions around stopping antibiotics may be aided by negative cultures at 36 hours, in children considered well.