A comparison of three rapid bacterial detection methods under simulated real-life conditions.

BACKGROUND: Bacterial screening of all produced platelet concentrates (PCs) is implemented in many countries to reduce the risk of transfusion-transmitted sepsis. This study compares three rapid bacterial detection methods by imitating real-life conditions. STUDY DESIGN AND METHODS: The sensitivity of a solid-phase scanning cytometer (optimized Scansystem, Hemosystem), fluorescence-activated cell sorting (FACS) analysis, and 16S RNA in-house nucleic acid testing (NAT) was evaluated by spiking PCs with four transfusion relevant bacteria (Staphylococcus aureus, Bacillus cereus, Klebsiella pneumoniae, and Escherichia coli ). Two different inocula (10 colony-forming units [CFUs]/mL and 10 CFUs/bag) were used to simulate real-life conditions. Samples were taken at 12, 16, 20, and 24 hours after spiking. RESULTS: With the high inoculum, NAT had a 100 percent rate of positive testing for all four types of bacteria (10/10 replicates) at each time point. With the exception of E. coli, the sensitivity of FACS and optimized Scansystem was comparable for the high inoculum. With the low inoculum, 60 percent of E. coli, 80 percent of B. cereus, 90 percent of K. pneumoniae, and 100 percent of S. aureus were NAT-positive 12 hours after spiking. In contrast, only 20 percent of E. coli, 10 percent of B. cereus, and 70 percent of K. pneumoniae were FACS-positive with the low inoculum 12 hours after spiking. CONCLUSIONS: In summary, the preliminary data revealed a higher sensitivity for NAT in comparison to FACS and optimized Scansystem under the defined study conditions. To imitate real-life conditions, further spiking studies with a low inoculum (10 CFUs/bag) and slower growing organisms should be conducted to examine the sensitivity of available detection systems.