Background: Blood culture results are frequently used to guide antibiotic decision-making, but culture contaminants need to be distinguished from true pathogens.
Aims: To assess the contamination rate of blood cultures and validate a method to distinguish between true bacteraemia and contamination.
Methods: We analysed blood culture results from children who were admitted to the paediatric ICU and paediatric wards at the Sardjito Hospital, Yogyakarta, Indonesia between December 2010 and February 2013. For each positive culture result, the type of isolated organism, time to positivity, and the number of positive culture sites were considered to classify the isolate as representing a true bacteraemia or contaminant.
Results: There were 1293 cultures obtained from blood and 308 (23.8%) were positive for bacterial growth. Fifty-three (4.1%) of the total cultures drawn fulfilled criteria for contaminants. The most common blood culture contaminants were coagulase-negative staphylococci.
Conclusion: Using standardized criteria, it is possible to implement a working method to identify true nosocomial infection from blood culture contaminant, and thus limit the effect of contaminated blood culture on irrational antibiotic use.