Background Over the last two decades, the residual risk of acquiring a transfusion-transmitted viral infection has been reduced to less than 1 : 1 000 000 via improvements in different techniques (e.g. donor selection, leuco-depletion, introduction of 3rd or 4th generation enzyme-linked immunosorbent assays and mini-pool nucleic acid testing (MP-NAT). In contrast, the risk for transfusion-associated bacterial infections has remained fairly stable, and is estimated to be in a range between 1 : 2000 and 1 : 3000. Platelets are at an especially higher risk for bacterial contamination, because they are stored at room temperature, which provides good culture conditions for a broad range of bacterial strains. To improve bacterial safety of blood products, different detection systems have been developed that can be divided into culture systems like BacT/ALERT or Pall eBDS, rapid detection systems like NAT systems, immunoassays and systems based on the FACS technique. Culture systems are used for routine bacterial screening of platelets in many countries, whereas rapid detection systems so far are mainly used in experimental spiking studies. Nevertheless, pathogen-reduction systems are currently available for platelet concentrates and plasma, and are under investigation for erythrocytes. Methods In this review, the functional principles of the different assays are described and discussed with regard to their analytical sensitivity, analytical specificity, diagnostic sensitivity, diagnostic specificity and clinical efficiency. The detection methods were clustered into three groups: (i) detection systems currently used for routine screening of blood products, (ii) experimental detection systems ready to use for routine screening of blood products, and (iii) new experimental detection systems that need to be investigated in additional spiking studies and clinical trials. Results A recent International Society of Blood Transfusion international forum reported on bacterial detection methods in 12 countries. Eight countries have implemented BacT/ALERT into blood donor screening, whereas in three countries only quality controls were done by culture methods. In one country; shelf-life was reduced to 3 days, so no bacterial screening was implemented. Screening data with culture methods can be used to investigate the prevalence of bacterial contamination in platelets. Differing results between the countries could be explained by different test definitions and different test strategies. Nevertheless, false-negative results causing severe transfusion-related septic reactions have been reported all over the world due to a residual risk of sample errors. Rapid screening systems NAT and FACS assays have improved over the last few years and are now ready to be implemented in routine screening. Non-specific amplification in NAT can be prevented by pre-treatment with Sau3AI, filtration of NAT reagents, or reduction of the number of polymerase chainreaction cycles. FACS systems offer easy fully automated handling and a handling time of only 5 min, which could be an option for re-testing day-5 platelets. New screening approaches like immunoassays, detection of bacterial adenosine triphosphate, or detection of esterase activity need to be investigated in additional studies. Conclusion Bacterial screening of blood products, especially platelets, can be done with a broad range of technologies. The ideal system should be able to detect one colony-forming unit per blood bag without a delay in the release process. Currently, we are far away from such an ideal screening system. Nevertheless, pathogen-inactivation systems are available, but a system for all blood components will not be expected in the next few years. Therefore, existing culture systems should be complemented by rapid systems like NAT or FACS especially for day-5 platelets.
