Abstract
BACKGROUND: Blood culture is essential in diagnosing blood stream infections (BSIs). It is well known that blood volume inoculated into the culture bottle is directly associated with blood culture positivity. However, effective interventions to encourage physicians to increase inoculum volume in pediatric patients are not known. Therefore, multifaceted interventions were performed to optimize the blood volume in our pediatric intensive care unit (PICU). METHODS: Prior to intervention, the recommended blood culture inoculum was greater than 1 ml for any child. The interventions were initiated in November, 2016. A simple guideline recommending blood volume according to body weight category (<10kg: 1ml, 10–20kg: 2ml, >20kg: 4ml for each bottle), was distributed by e-mail and placed on blood collection carts. Monthly feedback of blood volume collected and blood culture positivity was provided to all PICU medical staff who routinely perform blood sampling. Blood weight was determined by subtracting the initial bottle weight from the bottle weight on laboratory arrival. The blood volume was compared pre and post intervention. RESULTS: The blood volume submitted in 2,449 blood culture bottles was included in the analysis. Of these, 1,051 (42.9%) were in pre intervention and 1,398 (57.1%) were in post intervention period. The median blood volume and the percentage above recommended blood volume decreased slightly in body weight under 10 kg group (1.23 ml vs 1.13 ml, P < 0.001, 79.0% vs 73.6%, P = 0.016), but increased significantly in body weight between 10 and 20 kg group (1.42 vs. 1.79 mL, P = 0.003, 32.6% vs. 45.5%, P = 0.002) and over 20 kg group (1.60 mL vs 2.78 mL, P < 0.001, 16.4% vs. 32.4%, P = 0.001) (Figure 1). The blood culture was positive in 5.6% of the bottles that achieved recommended blood volume and 3.0% from the bottle that did not
(P = 0.003). CONCLUSION: The multifaceted intervention was effective in improving blood volume for culture in pediatric patients weighting over 10 kg. DISCLOSURES: All authors: No reported disclosures.