Bacterial contamination of whole-blood-derived platelets: the introduction of sample diversion and prestorage pooling with culture testing in the American Red Cross.

ABSTRACT

BACKGROUND: Bacterial sepsis following whole blood-derived platelet (WBP) transfusion has remained a substantial patient risk, primarily due to a lack of practical and effective means to limit or detect bacterial contamination. We describe the risk of reported septic reactions to WBPs and the introduction of prestorage-pooled whole blood-derived platelets (PSPs) collected using initial sample diversion and cultured for bacterial contamination. STUDY DESIGN AND METHODS: Product qualification and quality control (QC) testing with the Acrodose PL system (Pall Medical) were evaluated in four regional blood centers. Bacterial contamination risk was assessed by review of reported septic transfusion reactions to WBPs and by aerobic QC culture of leukoreduced PSPs utilizing automated microbial detection system cultures (BacT/ALERT 3D, bioMérieux). RESULTS: Before implementing PSPs (January 2003-December 2006), we distributed 2,535,043 WBP units and received 20 reports of septic reactions including 2 fatalities (7.9 per million [1126,752] reactions and 0.79 per million [11,267,522] fatalities). In October 2006, PSPs were effectively implemented with a product qualification success rate of 99.6 percent and a mean yield of 4.0 x 10(11) platelets (PLTs) per pool. Whole blood collection sets with sample diversion technology were introduced during the operational trial and decreased the rate of confirmed-positive bacterial culture of PSPs from 2111 (1474) to 965 (11036) per million (odds ratio, 0.46; 95% confidence interval, 0.22-0.95). No septic reactions to PSPs were reported (25,936 PSP units distributed). CONCLUSION: Sample diversion and bacterial culture are effective methods to reduce bacterial risk with WBP transfusion. Bacterial contamination of PSPs was assessed at 5.8-fold our current rate for apheresis PLTs utilizing comparable culture protocols.