Design of a Predictive Score for Contamination of Pediatric Blood Cultures

The contamination rate for blood cultures is high in pediatrics, due to different sampling techniques and the difficulty of sampling small-weight children, thus favoring contamination of the devices at the time of sampling. It is also more difficult to distinguish contamination from true bacteremia in children at an early stage, notably due to the limited number of vials that can be taken at any one time. On a daily basis, clinicians are faced with the choice of whether or not to initiate probabilistic antibiotic therapy when faced with the result of a positive blood culture, particularly when identification is not yet available, but only direct examination. Contamination has major consequences for patient management. Studies in adults have shown that contamination increases hospital length of stay by 4 to 5 days, laboratory costs by +20% and recourse to intravenous antibiotic therapy by +39%. In children, studies came to the same conclusion, with greater prescription of antibiotics, particularly intravenous antibiotics, in patients with contaminated blood cultures than in patients with sterile blood cultures. It also showed that 26% of patients with contaminated blood cultures were initially hospitalized because of the positivity of this test. The aim of this research is to determine the factors associated with contamination, in order to create a predictive score that would help clinicians in their decision-making when receiving the blood culture result as "positive".