Background: Blood culture is essential to diagnose sepsis. Recently, there have been progress of blood culture system and culture media for the high yield and early detection of microorganisms. VersaTREK system (Thermo Fisher Scientific, KS) detects g...
Background: Blood culture is essential to diagnose sepsis. Recently, there have been progress of blood culture system and culture media for the high yield and early detection of microorganisms. VersaTREK system (Thermo Fisher Scientific, KS) detects gas pressure change in the blood culture bottles when there is a growth of microorganisms. Resin bottles have been applied in the BacT/Alert 3D system (bioMerieux Inc., NC) recently. The author compared the performance of these two systems prospectively for the patients suspected of sepsis.
Methods: Among the blood cultures requested from June to December 2014 at the Gyeongsang National University Hospital, a total of 2,994 blood culture sets were analyzed. After collection of 20 ml from the peripheral vein, 5 ml of the blood were equally inoculated into aerobic bottle (REDOX1), an anaerobic (REDOX2) of VersaTREK System, an aerobic resin bottle (FA Plus), and an anaerobic resin bottle (FN Plus) of BacT/Alert 3D System. Positive rate, contamination rate, false negative rate, and time to detection (TTD) were analyzed for each bottle. TTD is defined as the interval between bottle entry to the positive signal in the machine. Medical records were reviewed to determine the clinical significance of the isolate. To know the false negative, all bottles with negative signal were subcultured after 5 days of incubation.
Results: REDOX1 showed positive in 280 (9.35%), REDOX2 in 184 (6.15%), FA Plus in 306 (10.22%), and FN Plus in 305 (10.19%), respectively. Combined with aerobic and anaerobic bottles, VersaTREK System showed positive in 299 sets (9.98%) and BacT/Alert 3D System in 371 sets (12.39%). Skin contaminants grew 55 (1.8%) in REDOX1, 36 (1.2%) in FN Plus, 31 (1.0%) in FA Plus, and 17 (0.6%) in REDOX2. After terminal subculture, growth was observed in 32 bottles in 24 sets (1.0%) among 2,332 sets showing signal-negative. They consisted of 2 REDOX1, 22 REDOX2, 1 FA Plus, and 8 FN Plus. Mean TTD of all pathogens was significantly shorter in FA plus than in REDOX1 (11.31 h vs. 14.00 h, P=0.015). Mean TTD for all pathogens was significantly shorter in FN plus than in REDOX2 (10.64 h vs. 26.61 h, P<0.001).
Conclusions: REDOX1 and REDOX2 of VersaTREK system was not efficient enough either to maximize pathogen recovery from blood culture or to detect pathogen early compared to FA Plus and FN Plus resin bottles of BacT/Alert 3D system. Especially the performance of REDOX2, the anaerobic media, was far below than the comparator, FN Plus resin media. Skin contamination rate and false negative rate were acceptable in both systems.