Abstract
Blood stream infection (BSI) in a child with cancer is a life-threatening condition while blood cultures (BCs) play a crucial role for the diagnosis. The current practice of obtaining a BC from all sources—peripheral vein, central venous catheter (CVC), all lumens—is controversial and therefore in this study, our aim was to evaluate the utility of obtaining BCs from all these available sources. A retrospective study of BC sets obtained from all newly diagnosed patients with malignancy was conducted. A total number of 633 BC sets from 123 boys and 88 girls (median age: 5 5/12 years) were evaluated during infection episodes from January 1, 2005, to August 31, 2010. Among these cases, 134 were classified as true BSI (21%), 468 as non-BSI, and 31 as false-positive. In 64 cases, the results from peripheral and CVC sources were discordant: 57 catheter positive–peripheral negative and seven catheter negative–peripheral positive. Consequently, seven out of 134 true BSIs (5.2%) would not have been identified if only a BC from a CVC had been obtained. Moreover, if no BCs from all lumens had been obtained we would have missed up to 25% of true-positive BSIs. In the same way, if BCs from all sources had not been drawn we would not have detected up to 52% of true-positive BSIs. Conclusion: Obtaining BCs during an infection episode from all sources in a child with cancer is still mandatory.