Inter-laboratory comparison of different rapid methods for the detection of bacterial contamination in platelet concentrates

Abstract

Background: Bacterial contamination of platelet concentrates still represents a major risk in transfusion medicine, and a variety of screening methods have been available to improve the safety of PCs. In the present study, the analytical quality of three different rapid screening methods (BactiFlow flow cytometry, Pan Genera Detection Assay, 23S rRNA RT-PCR) was evaluated in an inter-laboratory comparison in three different German blood services.

Methods: Samples were inoculated with different bacteria [Staphylococcus aureus, Staphylococcus epidermidis, Escherichia coli (two strains), Klebsiella pneumoniae (two strains), Enterobacter aerogenes (one strain), Serratia marcescens (one strain)] at different counts (4·5 × 10(3) -4·5 × 10(8) CFU/ml) alternating with negative samples in one transfusion facility. Samples were blinded with a random order for each screening method, shipped to partners and analysed immediately after receipt with different rapid screening methods.

Results: The inter-laboratory comparison revealed that the BactiFlow assay and 23S rRNA RT-PCR-screening detected all samples correctly (positive: 12/12, negative: 8/8). The Pan Genera Detection Assay test detected only four of the positive samples. Four of the non-detected positive samples were below the assay's detection limit. Another four inoculated samples with comparatively high bacteria counts were detected false negative (E. coli (two strains): 9·87 × 10(5) and 2·10 × 10(7) CFU/ml, respectively, K. pneumoniae: 4·79 × 10(6) CFU/ml, S. aureus: 6·03 × 10(5) CFU/ml). All rapid screening methods revealed no false-positive results.

Conclusions: Both BactiFlow and 23S rRNA RT-PCR demonstrated a high sensitivity to detecting bacterial contamination in PCs. The Pan Genera Detection Assay had some shortcomings regarding sensitivity, especially for the detection of Gram-negative strains.