BackgroundBlood culture contamination is associated with health care costs and potential patient harm. Diversion of the initial blood specimen reduces blood culture contamination. We report results of the "real-life" clinical implementation of this technique. MethodsFollowing an educational campaign, use of a dedicated diversion tube was recommended prior to all blood cultures. Blood culture sets taken from adults using a diversion tube were defined as "diversion sets," those without, "non-diversion" sets. Blood culture contamination and true positive rates were compared for diversion and nondiversion sets and to nondiversion historical controls. A secondary analysis investigated efficacy of diversion by patient age. ResultsOut of 20,107 blood culture sets drawn, the diversion group included 12,774 (60.5%) and the nondiversion group 8,333 (39.5%) sets. The historical control group included 32,472 sets. Comparing nondiversion to diversion, contamination decreased by 31% (5.5% [461/8333] to 3.8% [489/12744], P < .0001]. Contamination was also 12% lower in diversion than historical controls [3.8% (489/12744) vs 4.3% (1,396/33,174) P = .02)]. The rate of true bacteremia was similar. In older patients, contamination rate was higher, and the relative reduction associated with diversion decreased (54.3% amongst 20-40-year-olds vs 14.5% amongst >80-year-olds). ConclusionsUse of a diversion tube in the ED reduced blood culture contamination in this large real life observational study. Efficacy decreased with increasing age, which requires further investigation.