Sterility testing of platelet concentrates prepared from deliberately infected blood donations

BACKGROUND: In general the bacterial count in freshly donated blood is low and even lower in the corresponding platelet concentrates (PCs). By use of flow cytometry (FACS) for sterility testing, the reliability of early versus later sampling times was evaluated. STUDY DESIGN AND METHODS: Blood donations were spiked with various numbers of Staphylococcus epidermidis, Staphylococcus aureus, Bacillus cereus, and Klebsiella pneumoniae. The corresponding PCs were prepared by the buffy-coat method and stored at 22 degrees C. A 20-mL sample was collected from each PC directly after preparation and after 8 hours. Samples were stored at 35 degrees C. Sterility testing of both PCs and samples was by FACS analysis at different time points. RESULTS: All stored PCs were found positive by FACS analysis, with detection times ranging between 8 and 24 hours (K. pneumoniae, B. cereus), 8 and 91 hours (S. aureus), and 144 hours (S. epidermidis). In the samples incubated at 35 degrees C, bacteria were detected after 8 to 19 hours (K. pneumoniae, B. cereus), 8 to 67 hours (S. aureus), and 19 to 43 hours (S. epidermidis). Some of the samples did not contain bacteria. CONCLUSION: Detection times for slow-growing bacteria are significantly shortened when PC samples are incubated at 35 degrees C: the numbers of bacteria in freshly prepared PCs may, however, be so low that the samples drawn for sterility testing do not contain a single bacterium. Our results do not support a shortening of the 24-hour or greater sampling time recommended by the manufacturers of established test systems, because also for consistent detection by FACS, bacteria need to grow in the PCs to sufficient numbers.     

Evaluation of a new generation of culture bottle using an automated bacterial culture system for detecting nine common contaminating organisms found in platelet components

BACKGROUND: An automated bacterial culture system (BacT/ALERT 3D, bioMérieux) has been previously validated with a variety of bacteria in platelets. The recovery of bacteria in platelets using a new generation of culture bottles that do not require venting and that use a liquid emulsion sensor was studied. STUDY DESIGN AND METHODS: Bacillus cereus, Enterobacter cloacae, Escherichia coli, Klebsiella oxytoca, Staphylococcus aureus, Staphylococcus epidermidis, Serratia marcescens, Streptococcus viridans, and Propionibacterium acnes isolates were inoculated into Day 2 platelets to concentrations of 10 and 100 CFU per mL. Samples were then studied with current and new aerobic, anaerobic, and pediatric bottles. RESULTS: All organisms, except P. acnes, were detected in a mean time of 9.2 to 20.4 (10 CFU/mL) or 8.7 to 18.6 (100 CFU/mL) hours. P. acnes was detected in a mean time of 69.2 (10 CFU/mL) or 66.0 (100 CFU/mL) hours. The 10-fold increase in inoculum was associated with a mean 9.2 percent difference in detection time. The aerobic, anaerobic, and pediatric bottles had a mean difference in detection time (hours) between the current and new bottles of 0.10 (p=0.61), 0.4 (p=0.38), and 1.0 (p < 0.001), respectively. CONCLUSION: No difference in detection time between the current and new aerobic and anaerobic bottles was demonstrated. The new pediatric bottles had a small but significant delay in detection.     

Detection of bacteria in WBC-reduced PLT concentrates using percent oxygen as a marker for bacteria growth

BACKGROUND: The risk of receiving a PLT concentrate (PC) contaminated with bacteria may be 1000-fold greater than that of pathogenic viral transmission, yet surveillance for this risk is not generally practiced. A novel bacteria detection system (BDS) that overcomes the limitations of current systems is described. The BDS monitors percent oxygen (%O2) in air above aliquots of PCs that have been filtered to remove the confounding effect of respiring PLTs and residual WBCs. STUDY DESIGN AND METHODS: One-day-old WBC-reduced whole-blood-derived PCs (WBPCs) were inoculated with bacteria at 100 to 500 CFU per mL. After 30 minutes, 2- to 3-mL aliquots were processed through a PLT-reducing filter into a sample pouch containing sodium polyanethol sulfonate and entrained air. After incubation at 35 degrees C for at least 24 hours, the %O2 was measured within the pouch. Noninoculated WBC-reduced WBPCs (n = 155), confirmed free of bacteria by routine culture, were tested in a like manner. Results from the latter group of WBC-reduced WBPCs were used to distinguish contaminated from noncontaminated units. RESULTS: After a 24-hour incubation at 35 degrees C, 195 (96.5%) of the 202 sample pouches obtained from inoculated units were detected by the BDS. After an additional 6 hours at room temperature, those that remained and were tested were found positive. None of the noninoculated controls produced a positive reading. CONCLUSION: The BDS is easy to use and provides good levels of sensitivity and specificity.     

Evaluation of two detection methods of microorganisms in platelet concentrates

Background: The performance of a bacterial 16S ribosomal DNA real‐time polymerase chain reaction (PCR) assay was evaluated and validated with an automated culture system to determine its use for screening of platelet concentrates (PCs). Study Design and Methods: PCs were spiked with suspensions of Escherichia coli, Serratia marcescens, Staphylococcus epidermidis and St. aureus at 1, 10, and 100 colony‐forming units (CFUs) mL and stored for 5 days. DNA amplification was performed using real‐time PCR. The BacT/ALERT was used as a reference method and samples were inoculated into an aerobic culture bottle; for the PCR assay, aliquots were drawn from all (spiked) PCs on days 0 to 5 of storage. Results: Real‐time PCR detected only the gram‐positive bacteria in PCs spiked with low bacterial titres (1 CFU mL) after 48 h; however, it was able to detect all positive samples in PCs spiked with 10 CFU mL of either gram‐positive or gram‐negative bacteria after 48 h. In addition, real‐time PCR detected all positive samples in PCs spiked with high gram‐positive bacterial titres (100 CFU mL) after 24 h. On the other hand, the BacT/ALERT system showed positive results in all samples within 24 h. Conclusion: The BacT/ALERT method is more sensitive and should continue to be the gold standard for identifying bacterial contaminations in blood samples. The real‐time PCR approach can be used for the screening of PCs for microbial detection before they are released from blood centres or shortly before they are used in blood transfusion, and thus allow an extended shelf life of the platelets.     

Basics of flow cytometry-based sterility testing of platelet concentrates

BACKGROUND: Flow cytometry (FACS) is a common technique in blood banking. It is used, for example, for the enumeration of residual white blood cells in plasma and in cellular blood products. It was investigated whether it can also be applied for sterility testing of buffy coat-derived platelet concentrates (PCs). STUDY DESIGN AND METHODS: Plasma-reduced PCs were spiked with bacteria and stored at 20 to 24 or 37 degrees C for various times. The following 10 species were used: Bacillus cereus, Enterobacter cloacae, Escherichia coli, Klebsiella pneumoniae, Propionibacterium acnes, Pseudomonas aeruginosa, Staphylococcus aureus, Staphylococcus epidermidis, Serratia marcescens, and Yersinia enterocolitica. Bacterial DNA was stained with thiazole orange. After the platelets were lysed, bacteria were enumerated by FACS. RESULTS: All bacteria species used were detectable by FACS. The lower detection limit was approximately 100 bacteria per microL, that is, 10(5) per mL. In general, the titers measured were 1.2- to 3-fold higher than those determined by colony forming assay. In one case (K. pneumoniae) in which the dot plot of the bacteria cloud overlapped with that of bacteria debris, they were consistently lower. When PC samples were inoculated with approximately 1 colony-forming unit per mL of bacteria and kept at 37 degrees C, most species were detected within 21 hours or less. Exceptions were E. cloacae and P. acnes, which were detected after 24 to 40 and 64 hours, respectively. At 20 to 24 degrees C, the detection times were strongly prolonged. CONCLUSION: Sterility testing of PCs by FACS is a feasible approach. The present data suggest incubating PC samples for 20 to 24 hours at 37 degrees C before testing. For slow-growing bacteria, the incubation period must be prolonged by 1 to 2 days.     

First results using automated epifluorescence microscopyto detect Escherichia coli and Staphylococcus epidermidisin WBC-reduced platelet concentrates

BACKGROUND: Many methods have been tested for the detection of bacterial contamination in platelets. However, only those using molecular biology or cell culturing consistently detect contamination at levels below 10(5) bacteria per mL. This report describes the initial investigation into an alternative method that offers the possibilities of high sensitivity and rapid response while using available laboratory equipment and supplies. This method relies on a fluorescent nucleic acid stain, which preferentially stains bacteria but not platelets, and automated epifluorescence microscopy for rapid analysis. Measurements in WBC-reduced platelet concentrates (PCs) contaminated with bacteria are reported at concentrations between 10(3) and 10(6) bacteria per mL. STUDY DESIGN AND METHODS: Staphylococcus epidermidis or Escherichia coli was inoculated into aliquots of WBC-reduced PCs on Days 2 through 5 of storage. Bacterially inoculated and control PCs were stained, platelets and residual WBCs were lysed, and 200 microL of sample was filtered onto black polycarbonate filters. All preparations were done in triplicate. An automated epifluorescence microscope examined approximately 2 percent of the area of each filter and used image analysis to select the fluorescent particles that should be counted as bacteria. RESULTS: Samples containing 3 to 5 x 10(3) bacteria per mL produced about three times as many fluorescent particles classified as bacteria as the controls. Lower concentrations of S. epidermidis were detected because of higher fluorescence intensity. Simultaneous preparation of six samples requires about 35 minutes. Analysis of each prepared sample takes 10 minutes, for a total preparation and analysis time of about 95 minutes for 6 samples. CONCLUSION: Low concentrations (<5 x 10(3) bacteria/mL) of deliberately inoculated S. epidermidis or E. coli can be measured quickly in WBC-depleted PCs by using a fluorescent nucleic acid stain, differential lysis, and automated microscopy. Continued refinement of the method, studies employing other bacterial strains, and further validations of assay performance are warranted.     

16S rDNA实时荧光定量PCR法与全自动培养法在血小板细菌污染检测中的应用比较

目的比较16S rDNA实时荧光定量PCR法与全自动培养法在血小板制品细菌污染检测中的灵敏度和特异性,评价2种方法的应用前景。方法将血小板制品污染中常见的6种细菌用浓缩血小板悬液进行稀释,并选取浓度为102、101、100的菌悬液,分别用实时荧光定量PCR法和全自动培养法进行检测。结果用16S rDNA实时荧光定量PCR法对6株细菌检测,其灵敏度和特异性均为100%,Κ=1.000。用全自动培养仪对6株细菌检测,其特异性均为100%;灵敏度分别为:金黄色葡萄球菌需氧瓶95.5%,Κ=0.886,厌氧瓶90.9%,Κ=0.787;表皮葡萄球菌及大肠杆菌2种培养瓶均为83.3%,Κ=0.667;蜡样芽孢杆菌2种培养瓶均为86.7%,Κ=0.684;铜绿假单胞杆菌需氧瓶度为100%,Κ=1.000,厌氧瓶为44.4%,Κ=0.286;痤疮丙酸杆菌需氧瓶为16.7%,Κ=0.105,厌氧瓶为91.7%,Κ=0.886。结论实时荧光定量PCR法检测血小板制品中的细菌污染,灵敏度高,特异性好,且省时、经济,能应用于临床上血液样本的大规模筛查。     

Volume-reduced platelet concentrates: optimization of production and storage conditions

BACKGROUND: Plasma can be removed from platelet (PLT) concentrates (PCs) when volume reduction for PLT transfusion is indicated. Volume‐reduced PCs are currently produced from pooled buffy coat (BC) PCs or apheresis PCs by pretransfusion volume reduction, followed by transfer to a syringe for immediate transfusion. We evaluated the maximal storage time of the volume‐reduced PCs in gas‐permeable containers. STUDY DESIGN AND METHODS: Volume‐reduced PCs were produced from BC‐derived and apheresis PCs by hard‐spin centrifugation. Supernatant was removed and the PLTs were resuspended in 20 mL of retained original PC and had PLT concentrations ranging from 10.8 × 10⁹ to 13.8 × 10⁹ PLTs/mL. Volume‐reduced PCs were stored either in syringes or in containers made from diethylhexyl phthalate (DEHP)‐polyvinylchloride (PVC) or butyryl trihexyl citrate (BTHC)‐PVC plastic. Units were sampled at t = 0, 1, 3, and 6 hours for in vitro measurements. RESULTS: When prepared from 2‐day‐old PCs (n = 4), volume‐reduced PCs from BCs in a syringe had a pH_37°C of 5.76 ± 0.04 at t = 6 hours after volume reduction. In the DEHP‐PVC container, pH was 5.85 ± 0.15 (not significant), and in the BTHC‐PVC, 6.34 ± 0.16 (p < 0.001), at t = 6 hours. When made from 7‐day‐old PCs, pH was lower for all storage conditions: 5.68 ± 0.06 in the syringe, 5.70 ± 0.09 in the DEHP‐PVC container (not significant), and 6.07 ± 0.24 in the BTHC‐PVC container (p < 0.01) at t = 6 hours. Volume‐reduced 2‐day‐old apheresis PCs had a pH of 6.47 ± 0.20 at t = 6 hours. CONCLUSIONS: Adult‐dose PCs derived from BC or apheresis can be volume‐reduced to approximately 20 mL in a closed gas‐permeable system. Volume‐reduced PCs in BTHC‐PVC containers retain a mean pH of more than 6.0 up to 6 hours after production. Syringes allow only 3 hours of storage.     

Evaluation of an automated culture system for detecting bacterial contamination of platelets: an analysis with 15 contaminating organisms

BACKGROUND: Approximately 1 in 2000 platelet components are bacterially contaminated. The time to detection of 15 seeded organisms in platelets recovered from an automated culture system was studied. STUDY DESIGN AND METHODS: Isolates of Bacillus cereus, Bacillus subtilis, Candida albicans, Clostridium perfringens, Corynebacterium species, Enterobacter cloacae, Escherichia coli, Klebsiella oxytoca, Propionibacterium acnes, Pseudomonas aeruginosa, Staphylococcus aureus, Staphylococcus epidermidis, Serratia marcescens, Streptococcus pyogenes, and Streptococcus viridans were inoculated into Day 2 apheresis platelet components to obtain a final concentration of approximately 10 and 100 CFU per mL (2 units/organism). Each bag was sampled 10 times (20 mL/sample). Four mL of each sample was inoculated into standard aerobic and anaerobic bottles and into aerobic and anaerobic bottles containing charcoal; 2 mL was inoculated into pediatric aerobic bottles (so as to maintain a 1:10 ratio of sample to media) and 1 mL into thioglycollate broth. RESULTS: With the exception of P. acnes, all organisms were detected in a mean of 9.2 to 25.6 hours. A range of 10 serial dilutions in inoculating concentrations was associated with an overall 10.1-percent difference in detection time. A mean of 74.4 and 86.2 hours (100 and 10 CFU/mL inocula, respectively) was required for the detection of P. acnes in anaerobic bottles. CONCLUSION: Bacteria thought to be clinically significant platelet contaminants can be detected in 9.2 to 25.6 hours when the starting concentration is approximately 10 to 100 CFU per mL. P. acnes required considerably longer incubation times for detection (in either aerobic or anaerobic bottles). However, P. acnes is of questionable clinical significance. Such a detection system could be used in either a blood collection center or a transfusion service to screen platelet concentrates for bacterial contamination. Such testing (with sterile sampling performed so as to maintain a closed-bag system) would be expected to save lives and might allow an extension of platelet storage.     

A model to predict the improvement of automated blood culture bacterial detection by doubling platelet sample volume

BACKGROUND: Some transmissions of bacterial infection from platelet (PLT) transfusion have occurred since introduction of automated blood culturing of apheresis PLTs. The majority of these cases involved false-negative culture results in which slow-growing Staphylococcus organisms were implicated. This study analyses the projected benefit in detecting slow growing organisms by increasing sample volume from 4 to 8 mL. STUDY DESIGN AND METHODS: Bacterial growth was modeled by varying the initial inoculum, doubling time, and lag time. The numbers of organisms present at a 24-hour sampling time were calculated. Poisson analysis was used to determine the fraction of 4- or 8-mL samples that could be detected because they contained organisms. For each inoculum, the percentage of improved detection by doubling sample volume was defined as delta, the difference between the percentage of detection of 8- and 4-mL samples. RESULTS: The maximum improvement in detection by doubling sample volume was 25 percent and did not depend on bacterial growth rate or lag time. As inocula increased toward 40 colony-forming units per unit, delta decreased. As lag and doubling times increased, inocula corresponding to maximum delta increased and the width of the distribution curve broadened. CONCLUSION: Doubling sample volume will not double the chance of detecting slow growing organisms, but instead is predicted to improve detection by 25 percent or less.     

Operational feasibility of routine bacterial monitoring of platelets

Bacterial contamination of platelets poses the greatest risk of mortality and morbidity to platelet transfusion recipients. Some European countries have introduced routine bacterial monitoring of platelets to reduce the risk of transmission of bacteria. A pilot study was carried out at the Northern Ireland Blood Transfusion Service, using the BacT/ALERT automated culture system, to assess the operational feasibility of routine bacterial monitoring of platelets. About 4885 platelet concentrates (PCs) were tested in a 1-year period. Of the 28 (0.57%) initial reactive cultures, 13 (46%) were reproducible on repeat culturing. Of these, 10 were detected within 24 h of incubation either in aerobic or both aerobic and anaerobic culture bottles. A sample of time-expired units (423) that had initial negative culture results remained negative when retested on day 8. About 213 time-expired units were subjected to routine quality assessment and more than 85% were found to conform to quality standards specified in the UKBTS guidelines for platelet count (> or =240 x 10(9) per adult dose PC) and pH (6.4-7.4). There was a reduction in the platelet count because of the volume removed (15 mL) for sampling. Routine bacterial testing with day 2 sampling and a negative culture result after 24 h as a mandatory release criterion would improve product safety. Implementation of 100% testing would be operationally feasible but may require extension of the shelf life if unacceptable wastage is to be avoided.     

Culture-based bacterial detection systems for platelets: the effect of time prior to sampling and duration of incubation required for detection with aerobic culture

BACKGROUND: Bacterial contamination of platelet (PLT) products occurs at low concentrations requiring a period of incubation for growth to minimize sampling error. Culture-based detection methods also need sufficient incubation time; together these periods may limit the useful life of PLTs. This study characterizes the impact of sampling and detection times with two commercially available bacteria detection products. STUDY DESIGN AND METHODS: Apheresis PLTs inoculated with nine bacterial species at low concentrations were sampled immediately and 24 hours after inoculation. Test results were analyzed after incubation at 16, 20, and 24 hours after sampling with two bacterial detection systems. RESULTS: When sampled immediately after inoculation, two commercially available bacterial detection systems (BacT/ALERT, bioMérieux; and eBDS, Pall Corp.) failed to detect some PLTs inoculated with Staphylococcus epidermidis, Serratia liquefaciens, or Pseudomonas aeruginosa and S. epidermidis, S. liquefaciens, Bacillus cereus, or P. aeruginosa, respectively. The BacT/ALERT was better at 20 hours (p<0.02), but not at 16 or 24 hours for Time 0 sampling. When sampling occurred 24 hours after inoculation, there were no difference between the two systems. CONCLUSION: Results suggest that for either bacteria detection system, holding PLTs for 24 hours before sampling improves the detection sensitivity for PLTs contaminated with low concentrations of bacteria, and longer incubation periods improve detection.     

Multiple pH measurement during storage may detect bacterially contaminated platelet concentrates

BACKGROUND: Bacterial contamination or platelet (PLT) metabolism can change the pH of stored PLT concentrates (PCs). Measurement of pH for quality control is currently done on a limited basis. An easy noninvasive method was developed to obtain sequential pH measurements over time, without risking contamination and/or consuming PCs. STUDY DESIGN AND METHODS: The objective was to measure pH profiles of bacterially contaminated PCs over 7 days of storage. Small‐volume PC storage bags with incorporated pH sensor were prepared and in vitro variables were tested using aliquots of PCs. The pH sensors were used to delineate trends associated with the deterioration of these PCs upon inoculation with 19 different bacterial strains and one yeast. RESULTS: Monitoring the pH trends in real time in a noninvasive fashion, most bacterial strains were detected within 24 to 72 hours after spiking into the bag. At the time of detection, bacterial concentrations had reached levels between 1 × 10³ and 1 × 10⁸ colony‐forming units/mL. Several strains had pH rebound after initial drop. Multiple noninvasive pH reads allowed bacterial detection whereas single pH reads could give false‐negative results. CONCLUSIONS: The noninvasive pH sensor facilitated the detection of most strains of bacterial contaminants within 3 days with no potential for sampling error.     

Elimination and multiplication of bacteria during preparation and storage of buffy coat-derived platelet concentrates

BACKGROUND: The prevalence of bacterial contamination of random-donor platelet concentrates (PCs) is considerably lower than that of blood donations. Which key steps of the preparation procedure contribute to the elimination of bacteria was investigated. STUDY DESIGN AND METHODS: Ten bacteria species were used. Blood donations were spiked with bacteria and stored at 22 degrees C for 8 hours. The buffy coats were kept for 6 hours. PCs were prepared from pools of 4 buffy coats. At each preparation step and during PC storage, bacteria contents were measured. In additional experiments, the titers of spiked blood and buffy coats were determined after storage at 20, 22, or 24 degrees C for 8 and up to 24 hours, respectively. RESULTS: Enterobacter cloacae, Escherichia coli, Pseudomonas aeruginosa, Serratia marcescens, and Yersinia enterocolitica were completely inactivated during storage in blood or buffy coats. Titer reduction was between 3.32 and 4.62 log. Bacillus cereus, Propionibacterium acnes, Staphylococcus aureus, and Staphylococcus epidermidis did not multiply. Compared with their values in spiked blood the titers in the PCs were reduced by 1.7 to 2.8 log. Klebsiella pneumoniae was the only species that grew in blood. With the exception of P. acnes, those species that were not removed by the preparation process multiplied in the PCs. Remarkable donor-to-donor variations of the bactericidal activities of buffy coats were detected when the storage time was prolonged to 24 hours. CONCLUSIONS: Bacteria are significantly eliminated by the preparation procedure for random donor PCs. Also, blood and buffy coats are bactericidal for most species. When buffy-coat storage is prolonged, it cannot, however, be predicted whether specific strains vanish or multiply.     

Rapid identification of bacterially contaminated platelets using reagent strips: glucose and pH analysis as markers of bacterial metabolism

BACKGROUND: One in every 1000 units of platelets is bacterially contaminated, which puts patients at risk for transfusion-associated sepsis and death. However, there is currently no screening test in place to detect contaminated units. The use of commercially available multiple-reagent urine dipsticks for this purpose was evaluated. STUDY DESIGN AND METHODS: Platelet concentrates were inoculated with either sterile saline or suspensions of Staphylococcus aureus, Staphylococcus epidermidis, Bacillus cereus, Klebsiella pneumoniae, or Serratia marcescens to a final concentration of 50 colony-forming units (CFU) per mL. The platelets were analyzed daily by the use of multiple- reagent strips, quantitative culture, and glucometry. RESULTS: B cereus grew rapidly, reaching 10(7) CFU per mL 1 day after inoculation, while S. epidermidis grew slowly, achieving similar concentration 4 to 6 days after inoculation. Two of 10 dipstick reagents, glucose and pH, proved useful in detecting bacteria. Both were lower in bacterially contaminated units than in controls. Glucose data obtained from automated analyzers validated the dipstick data. All organisms were detected at concentrations > or = 10(7) CFU per mL, and S. aureus and K. pneumoniae were detected in the range of 10(3) to 10(5) CFU per mL. CONCLUSION: The multiple-reagent test used had a sensitivity and specificity of 95 percent (> or = 10(7) CFU/mL) and 98 to 100 percent, respectively. These data indicate that urine dipsticks can be used to rapidly and inexpensively detect bacterial contamination in platelet concentrates, which potentially will reduce morbidity and mortality at minimal cost.     

Comparison of bacteria growth in single and pooled platelet concentrates after deliberate inoculation and storage

BACKGROUND: The ability to store pools of platelet concentrates (PCs) for extended periods would provide logistical flexibility. However, reports of severe adverse reactions due to the transfusion of contaminated PCs led to an examination of whether the total bacteria levels after storage of pools containing a deliberately inoculated platelet unit would be significantly different than the levels in paired unpooled concentrates. STUDY DESIGN AND METHODS: A single PC was deliberately inoculated on Day 0 with one of three bacterial species (0.1–8.0 colony-forming units/mL). On Day 1, the deliberately inoculated PC was divided into three equal parts and either 1) pooled with 5 half-volume, ABO- and Rh-identical PCs; 2) similarly pooled and white cell reduced; or 3) kept as a control. Sterile connections were used during pooling; modified storage containers were used to ensure the correct surface-to-volume ratio of the single unit. RESULTS: Between Day 2 and Day 5 of storage, in 26 of 36 paired samples, nonfiltered pools containing Escherichia coli had greater total numbers of bacteria than did the paired single PCs. Day 2 pools had total bacteria levels approximately five times higher (colony-forming units/mL × container volume) than those in single units (p < 0.05). There was rapid growth of Staphylococcus aureus by Day 2 in pooled and unpooled PCs; by Day 3, total bacteria levels were approximately five times higher in pools than in single units (p < 0.05). Between Days 3 and 5 of storage, in 23 of 27 paired samples, nonfiltered pools containing S. aureus had greater total bacteria levels than the single PCs. By Day 5, 15 of 16 non-white-cell reduced pools had total levels of Staphylococcus epidermidis bacteria approximately five times those in the paired single PCs. Greater total bacteria levels in pooled units than in single units generally occurred when bacteria in pools reached the stationary phase of growth (when bacteria concentration became constant), and they were well correlated with the sixfold volume of pooled units. White cell reduction did not substantially affect the time required to attain stationary phase. CONCLUSION: The potential during storage for greater total bacteria levels in pools than in single PCs is a consequence of the greater volume of the pool.     

Value of anaerobic culture in bacterial surveillance program for platelet concentrates

BACKGROUND: Short‐term aerobic bacterial culture (STABC) has been used routinely in Hong Kong since 1998 to reduce bacterial contamination in platelet concentrates (PCs) with good results. With more countries implementing routine aerobic and anaerobic cultures of PCs, a prospective study was conducted to determine the value of anaerobic culture to STABC. STUDY DESIGN AND METHODS: PC tested by STABC was used as control. Twenty milliliters of the PC selected for this study was aliquoted and pooled for 7 days aerobic and anaerobic culture. If the initial culture was positive, samples retrieved from the original PC and their associated components were cultured for confirmation and microbiologic identification. RESULTS: A total of 10,035 PC units (2007 pools) were tested. The confirmed positive rates by aerobic and anaerobic cultures per pool were 3 (0.15%) and 13 (0.65%), respectively, which was equivalent to an increased yield from 0.03 to 0.13 percent of PC if anaerobic culture was added. Of the 10 bacteria detected by anaerobic culture only, 9 were found to be Propionibacterium acnes and the remaining one Peptostreptococcus sp. Their mean detection time from inoculation was 92.16 hours (range, 50.4‐124.8 hr). CONCLUSION: Addition of anaerobic culture to our routine STABC would significantly increase the detection rate of bacterial contaminated PC. However, since only slow‐growing bacteria were detected, and because their clinical significance was uncertain, it is concluded that there was no clear justification to introduce anaerobic culture locally if 5‐day shelf life for PCs was to be maintained.