1997. Impact of Blood Culture Fill Volumes

Abstract

BackgroundHistorically, increases in blood culture (BC) fill volumes (FVs) have been shown to increase yield of BCs and lower contamination rates. Low FV are a common cause of false negative BCs. 10 mL is considered an ideal FV for a BC. In 2015 and 2016, at North Shore University Hospital, FVs averaged <5 mL per BC. In 2017, several interventions were implemented to increase FVs, including convening informal meetings and seminars to educate nursing staff, educational phlebotomy posters, placing 10-mL markings on BC bottles and using butterfly catheters and tabletops for collection. Our aim was to assess trends in overall yield (OY), contaminants and FVs.MethodsAverage FVs, positive BC quantities and organism identification were obtained from 2015 through 2017. Contaminants included bacillus, corynebacterium, coagulase negative staphylococcus, micrococcus and single sets of α-hemolytic streptococcus. OY was the number of positive sets, excluding contaminants, divided by the total number of BCs. Subgroup yield (SY) was the number of positive sets in a subgroup divided by the total number of BCs. Trends in OY, SY, and contaminants were assessed using the Cochran Armitage Trend test. The one-way ANOVA test was used to assess differences between FVs by year.ResultsOY increased over the 2015–2017 period (Table 1; P < 0.0001). All SYs increased except for staphylococcus and anaerobes. Contaminants did not show a decreasing trend (Table 2; P = 0.9002).Table 1.Yield by YearYield, %Year P Value201520162017OY6.377.578.49<.0001Total Gram-positivesa3.13.523.69<.0001Staphylococcus1.792.011.950.1178Enterococcus0.660.880.98<.0001Gram-negative2.523.033.64<.0001Anaerobe0.220.240.260.3811Fungal0.150.280.270.001Mixed0.280.380.420.0026Other0.090.120.21<.0001 aIncludes staphylococcus and enterococcus.Table 2.Mean FV by YearYearTotal BC ordered, nContaminants, %Mean FV (SD), mL201535,6242.874.32 (0.26)201638,4402.994.39 (1.39)201737,0422.886.11 (0.98)ConclusionBC FVs successfully increased with interventions implemented. While OY increased each year, an association between yield and FVs could not be determined due to small sample sizes. Further evaluation at quarterly intervals is ongoing and may help establish a correlation.Disclosures All authors: No reported disclosures.