Residual risk of bacterial contamination of platelets: six years of experience with sterility testing

Abstract

BACKGROUND Canadian Blood Services screens 100% of platelet concentrates (PCs) for bacterial contamination with the BacT/ALERT system. Quality‐control sterility testing of 1% (≥10 units) of outdated PCs is performed monthly. Data from routine screening, quality‐control testing, and septic reactions obtained from 2010 to 2016 are presented herein. STUDY DESIGN AND METHODS In total, 601,988 buffy coat PC pools and 186,737 apheresis PCs were routinely screened with aerobic cultures over 6 years. Outdate quality‐control testing of 8535 buffy coat and 8498 apheresis PCs was performed using aerobic and anaerobic cultures during the same period. Results were classified as “true‐positives” when the same bacterium was isolated in initial and confirmatory cultures or “false‐negatives” when bacteria were missed in early screening and were captured during quality‐control sterility testing or through investigation of sepsis cases. RESULTS During routine screening, the true‐positive rates between buffy coat (0.94 per 10,000) and apheresis (0.96 per 10,000) PCs were similar (p = 0.9473). Seventy‐five bacteria isolated during PC screening included Gram‐positive and Gram‐negative organisms. Six false‐negative septic reactions were reported that implicated coagulase‐negative staphylococci (n = 3) and Staphylococcus aureus (n = 3) for approximate rates of 1 per 100,000 transfusion reactions and 1 per 500,000 fatalities. During quality‐control testing, the false‐negative rates between buffy coat (8 per 10,000) and apheresis (9 per 10,000) PCs were similar (p = 0.7897). All 15 quality‐control isolates were Gram‐positive bacteria. CONCLUSION The current bacterial screening protocol is efficacious for identifying Gram‐negative bacteria. However, the high proportion of Gram‐positive organisms detected on outdate quality‐control testing and septic transfusion events demonstrates a residual safety risk that merits further intervention.