Microbial Contamination in Cord Blood (CB): A Retrospective Analysis of Three Test Protocols

Abstract

Aim Since 1995, the SCBB has used three protocols to screen CB for contaminating microbes. The protocols differed in the number of culture bottles inoculated and sample volume used. The aim of this study was to compare microbial detection rates for each method. Method CB units (n = 9,618) were tested for microbial contamination using BacT/ALERT blood culture bottles (BioMerieux). Of these, 6,836 had 20 ml plasma waste fraction inoculated into adult anaerobic and aerobic bottles (Group A), 365 had 1 ml final stem cell product into an aerobic paediatric bottle (Group B), and 2,417 had 1 ml final product into a paediatric bottle and 10 ml plasma waste fraction into an adult anaerobic bottle (Group C). All bottles were incubated for a minimum of 5 days. Detection rates for microbe contamination were calculated for each group. Result Microbe detection rates were: Group A –345/6836 (5.0%), Group B – 3/365 (0.8%) and Group C – 66/2417 (2.7%)(p < 0.0001). Organisms detected by A but not by B (and to a lesser extent C) included Streptococcus, Staphylococcus, Bacteroides and Proprionibacterium. Conclusion Contamination detection rates differ significantly, depending upon the sample volume and type of culture bottle used. A protocol that uses the recommended sample volume and both anaerobic and anaerobic test bottles is likely to detect the highest number of contaminated units. Testing of a small sample from the final product using only a single paediatric aerobic bottle is likely to miss detection of the majority of anaerobic and some aerobic organisms. Prospective studies to define a standardised method for microbial detection for CB are required.