Sepsis associated with blood transfusion

Clinical significance of bacteriologic screening in platelet concentrates

BACKGROUND

Despite routine bacterial screening with a bacterial culturing system (BacT/ALERT, bioMerieux) of platelet (PLT) concentrates, two cases of life-threatening sepsis attributed to transfused PLT products contaminated with Bacillus cereus were reported to the regional hemovigilance office in the southwest region of the Netherlands. These reports necessitated a retrospective evaluation of the currently applied bacteriologic screening program.

STUDY DESIGN AND METHODS

Bacteriologic screening of all PLT concentrates on production was introduced in October 2001. Aliquots (5-10 mL) of pooled PLT concentrates in additive solution were taken for cultures with the BacT/ALERT system 14 to 24 hours after donation. The total culturing period was 7 days and in case of positive signals, identification cultures were taken from the culture bottles. The results of the bacterial screening, identification, and clinical significance of possibly contaminated pooled PLT concentrates were evaluated retrospectively over a 2-year period.

RESULTS

In this period, a total of 28,104 pooled PLT concentrates were produced. Positive bacterial screening was found in 0.72 percent (n = 203). Of these, in 184 pooled PLT concentrates bacteria were cultured and identified, and in the remaining 19 (9.4%) identification cultures were negative. Before a positive screening was found, 113 PLT concentrates had already been transfused without the occurrence of clinical significant transfusion reactions.

CONCLUSION

Bacterial contamination of pooled PLT concentrates was not related to clinically significant transfusion reactions. Despite negative screening for bacterial contamination, life-threatening transfusion-transmitted infections by contaminated PLT products can still occur. Other strategies should be applied to guarantee a higher degree of bacteriologic safety.

Impact of donor arm skin disinfection on the bacterial contamination rate of platelet concentrates

BACKGROUND AND OBJECTIVES

Despite improved methods for detecting bacterial contamination of blood products, bacterial sepsis remains a significant risk in blood transfusion. This study was undertaken to investigate whether adopting a different skin disinfection protocol could reduce the rate of bacterial contamination of platelet concentrates.

MATERIALS AND METHODS

Two skin disinfection protocols were consecutively used in the routine blood collection setting during two 10-month periods: 0.5% cetrimide/0.05% chlorhexidine solution followed by 70% isopropyl alcohol (first 10-month time-period); and 10% povidone-iodine followed by 70% isopropyl alcohol (second 10-month time-period). The rates of bacterial contamination of platelet concentrates were monitored by using a surveillance programme described previously.

RESULTS

The overall bacterial contamination rate in the first time-period was 0.072%. After introduction of the povidone-iodine and isopropyl alcohol protocol, the bacterial contamination rate decreased to 0.042% (relative risk reduction: -0.42; 95% confidence interval, -0.12 to -0.61, P= 0.009). There were no differences in the types of micro-organisms identified (P = 0.7).

CONCLUSIONS

Skin disinfection by povidone-iodine and isopropyl alcohol is more effective than that by cetrimide/chorhexidine and isopropyl alcohol in reducing venepuncture-associated contamination of platelet concentrates by skin flora. Our data indicate that the disinfection protocol should be used on a routine basis and such implementation should translate into a significant improvement in blood safety to patients receiving platelet transfusion.

Evaluation of the BacT/Alert automated blood culture system for detecting bacteria and measuring their growth kinetics in leucodepleted and non-leucodepleted platelet concentrates

BACKGROUND AND OBJECTIVE

To evaluate the BacT/Alert automated blood culture system for the detection of bacteria in platelet concentrates, and to determine bacterial growth kinetics in leucodepleted and non-leucodepleted units.

MATERIALS AND METHODS

Apheresis (Cobe Leucocyte Reduction System [LRS]) and pooled buffy coat-derived (Optipress) platelet concentrates (PCs) were tested. Six organisms were used for spiking the PCs: Clostridium perfringens, Bacillus cereus, Group B Streptococcus, Staphylococcus epidermidis, Staphylococcus aureus and Escherichia coli. Units were inoculated to give a final concentration of approximately equal to 1 and 50 colony-forming units (CFU)/ml. On days 0, 2 and 5, BacT/Alert standard aerobic and anaerobic bottles were inoculated with a 5-ml fill volume and bacteria were enumerated.

RESULTS

The BacT/Alert Automated blood culture system gave rapid determination times of spiked units, with all positives detected within 48 h and 98.1% detected within 24 h. In general, as the inoculum concentration increased, the detection time decreased. Rapid growth was obtained with all organisms tested except for B. cereus, which failed to grow on four occasions. Bacterial numbers on day 2 ranged from 10(5) to 10(11) CFU/ml and on day 5 ranged from 10(4) to 10(12) CFU/ml. Growth was not significantly greater in leucodepleted units.

CONCLUSIONS

The study confirmed that PCs are an excellent growth medium for bacteria. Rapid and substantial growth was obtained with all organisms under test. Leucodepletion does not appear to enhance bacterial proliferation. The BacT/Alert automated blood culture system could rapidly detect contamination of units. Bacterial screening using an automated blood culture system is therefore a potential option.

[Blood transfusion and bacterial risk]

Initial hemovigilance data confirm the incidence and severity of transfusion reactions due to the bacterial contamination of blood components (TRBC). With 18 deaths reported through the French hemovigilance network over the past 5 years, bacterial risks represent one of the major immediate complications of blood components (BC) transfusion. BC contamination may lead to more or less severe TRBC, depending on their origin: bacteria growth, the BC itself or unknown origin. Although the rate of donated blood or BC contamination is known (0.5% and 0.05%, respectively) it is still difficult to assess the actual incidence of TRBC, as it is difficult to identify them and relate them to transfusion. Likewise, better knowledge of bacteria, symptoms and outcome is required to improve prevention methods. Better prevention can reduce BC contamination and proliferation of bacteria at each stage of blood transfusion. Methods to detect BC contamination are still under investigation. Through continuous education of hemovigilance actors in identifying and dealing with TRBC, as well as drawing up procedures to perform inquiries and specific bacterial analyses, case reporting can be further improved in order to achieve more efficient prevention.

[Blood transfusion and bacterial risk]

Initial hemovigilance data confirm the incidence and severity of transfusion reactions due to the bacterial contamination of blood components (TRBCs). With 18 deaths reported through the French hemovigilance network over the past five years, bacterial risks represent one of the major immediate complications of BC transfusion. BC contamination may lead to more or less severe TRBCs, depending on their origin: bacteria growth, the BC itself or unknown origin. Although the rate of donated blood or BC contamination is known (0.5% and 0.05%, respectively), it is still difficult to assess the actual incidence of TRBCs, as it is difficult to identify and relate them to transfusion. Likewise, a better knowledge of bacteria, symptoms, and outcome is required to improve prevention methods. Better prevention can reduce BC contamination and proliferation of bacteria at each stage of blood transfusion. Methods of detecting BC contamination are still under investigation. Through continuous education of hemovigilance participants in identifying and dealing with TRBCs, as well as drawing up procedures to perform inquiries and specific bacterial analyses, case reporting can be further improved, in order to achieve more efficient prevention.

Bacillus cereus causing fulminant sepsis and hemolysis in two patients with acute leukemia

PURPOSE

Hemolysis is so rarely associated with Bacillus cereus sepsis that only two very well documented cases have been reported. This article reports two unusual cases of Bacillus cereus sepsis with massive intravascular hemolysis in patients who had acute lymphoblastic leukemia (ALL).

PATIENTS AND METHODS

A 20-year-old woman who was 9 weeks pregnant experienced a relapse of ALL. A therapeutic abortion was performed. During week 4 of reinduction the patient had abdominal pain, nausea, and vomiting, with severe neutropenia but no fever. Her condition deteriorated rapidly with cardiovascular collapse, acute massive intravascular hemolysis, and death within hours of the onset of symptoms. Blood cultures were positive for Bacillus cereus. Postmortem histologic examination and cultures revealed Bacillus cereus and Candida albicans in multiple organs. The second patient, a 10-year-old girl, presented with relapsed T-cell ALL. In the second week of reinduction, she had abdominal pain followed by hypotension. Again, no fever was noted. Laboratory studies showed intravascular hemolysis 12 hours after admission. Aggressive support was promptly initiated. Despite disseminated intravascular coagulation; cardiovascular, hepatic, and renal failure; and multiple intracerebral hypodense lesions believed to be infarcts, the patient recovered fully and resumed reinduction therapy.

CONCLUSIONS

Bacillus cereus infection can have a fulminant clinical course that may be complicated by massive intravascular hemolysis. This pathogen should be suspected in immunosuppressed patients who experience gastrointestinal symptoms and should not be precluded by the absence of fever, especially if steroids such as dexamethasone are being given. Exchange transfusion may be lifesaving in Bacillus cereus septicemia associated with massive hemolysis.

[Quantitative bacteriological evaluation of a method for skin disinfection in blood donors]

Skin disinfection at the site of venipuncture is a critical point in every blood transfusion collection procedure, as it contributes to ensure the bacterial safety of transfusion. Quantitative and qualitative analysis of bacteria present in the antecubital fossae before and after skin disinfection may be one method of assessing the anti-bacterial efficiency of disinfection. Swab culture systems and contact plates are the two techniques usually employed for this purpose. A washing and swabbing technique was used to quantify bacteria before and skin disinfection of the antecubital fossae in blood donors. This contra-placebo study was carried out on 32 donors, each of whom served as his own control, with a random choice of test arm and opposing control arm. Bacterial counts were determined in the antecubital fossae without skin disinfection (control, n = 32) and after a 3 step skin preparation procedure (cleaning, wiping, disinfection) using placebo (distilled water, n = 16) or an antiseptic product (mixture of chlorexidine, benzalkonium chloride and benzylic alcohol, n = 16). The absence of a statistical difference in bacterial counts between the right and left antecubital fossae without disinfection was controlled in a preliminary study of 20 subjects. Mean bacterial counts were 25,000/cm2 and 27,400/cm2 respectively for aerobic and anaerobic bacteria before disinfection, with a wide variation in results between individuals. When using placebo, preparation of the venipuncture site by the 3 step method (cleaning, wiping, disinfection) resulted in a non significant mean reduction of 0.56 log in aerobic and anaerobic bacteria. Using the antiseptic product, the same method resulted in a significant mean reduction of 1.8 and 1.7 log respectively in aerobic (p = 0.015) and anaerobic flora (p = 0.005). On an average, 2,750 aerobic bacteria/cm2 and 2,910 anaerobic bacteria/cm2 remained after disinfection, while qualitative analysis showed that disinfection suppressed the transitory flora in all cases but left part of the resident flora in 12/16 cases. These findings are comparable to those of other studies carried out to evaluate this kind of technique for the disinfection of operation sites. In comparison with other techniques classically employed for this type of evaluation (swab systems or contact plates), the method used in this study was the advantage of allowing the quantification of the reduction in bacteria. Hence this method could be employed for comparative assessment of skin disinfection techniques with the aim of improving their anti-bacterial efficiency and could also make possible the definition of a minimum bacterial count (resident flora) to be obtained in all cases after disinfection.