BACKGROUND: Bacterial contamination of platelet components (PCs) remains an important cause of transfusion-associated infectious risk. In 2004, Canadian Blood Services (CBS) implemented bacterial testing of PCs using the BacT/ALERT 3D system (bioMérieux). This system has been validated and implemented and continuous monitoring of culture rates allows gathering of data regarding true and false positives as well as false negatives. STUDY DESIGN AND METHODS: National data gathered between March 2004 and October 2010 from 12 CBS sites were analyzed to compare bacterial contamination rates across three platelet (PLT) preparation methods: apheresis, buffy coat, and PLT-rich plasma. Data were compared before and after implementation of protocol changes that may affect bacterial detection or contamination rates. RESULTS: Initial positive rates among the three production methods were significantly different, with apheresis PCs being the highest. The rates of confirmed positives among production methods did not differ significantly (p = 0.668). Increasing sample testing volumes from 4 to 6 mL to 8 to 10 mL significantly increased the rate of initial positives, while confirmed positives increased from 0.64 to 1.63 per 10,000, approaching significance (p = 0.055). Changing the skin disinfection method from a two-step to a one-step protocol did not significantly alter the rate of confirmed positives. During the period of data analysis, eight false-negative cases were reported, with five implicated in adverse transfusion reactions. CONCLUSION: Bacterial testing of PCs and implementation of improved protocols are incrementally effective in reducing the risk of transfusion of bacterially contaminated PLT concentrates; however, the continued occurrence of false-negative results means the risk has not been eliminated.
BACKGROUND: Bacterial contamination of platelet components (PCs) remains an important cause of transfusion-associated infectious risk. In 2004, Canadian Blood Services (CBS) implemented bacterial testing of PCs using the BacT/ALERT 3D system (bioMérieux). This system has been validated and implemented and continuous monitoring of culture rates allows gathering of data regarding true and false positives as well as false negatives. STUDY DESIGN AND METHODS: National data gathered between March 2004 and October 2010 from 12 CBS sites were analyzed to compare bacterial contamination rates across three platelet (PLT) preparation methods: apheresis, buffy coat, and PLT-rich plasma. Data were compared before and after implementation of protocol changes that may affect bacterial detection or contamination rates. RESULTS: Initial positive rates among the three production methods were significantly different, with apheresis PCs being the highest. The rates of confirmed positives among production methods did not differ significantly (p = 0.668). Increasing sample testing volumes from 4 to 6 mL to 8 to 10 mL significantly increased the rate of initial positives, while confirmed positives increased from 0.64 to 1.63 per 10,000, approaching significance (p = 0.055). Changing the skin disinfection method from a two-step to a one-step protocol did not significantly alter the rate of confirmed positives. During the period of data analysis, eight false-negative cases were reported, with five implicated in adverse transfusion reactions. CONCLUSION: Bacterial testing of PCs and implementation of improved protocols are incrementally effective in reducing the risk of transfusion of bacterially contaminated PLT concentrates; however, the continued occurrence of false-negative results means the risk has not been eliminated.
Authors: Seema Kacker; Evan M Bloch; Paul M Ness; Eric A Gehrie; Christi E Marshall; Parvez M Lokhandwala; Aaron A R Tobian Journal: Transfusion Date: 2019-01-08 Impact factor: 3.157
Authors: Burak Bahar; Wade L Schulz; Amit Gokhale; Bryan R Spencer; Eric A Gehrie; Edward L Snyder Journal: Br J Haematol Date: 2019-09-30 Impact factor: 6.998
Authors: Maria Loza-Correa; Juan A Ayala; Iris Perelman; Keith Hubbard; Miloslav Kalab; Qi-Long Yi; Mariam Taha; Miguel A de Pedro; Sandra Ramirez-Arcos Journal: PLoS One Date: 2019-01-25 Impact factor: 3.240