10 Years Experience with Bacterial Screening of Platelet Concentrates in the Netherlands

Use of flow cytometry method to detect contaminations of platelet suspensions

In this study, it was aimed to investigate bacterial contamination in apheresis platelet suspensions (APS) by automated blood culture system and flow cytometry method (FCM).33 spiked APS each using 11 bacterial strains (5 standard strains, 6 clinical isolates), were prepared in three different dilutions (1-10, 10-50, 50-100 cfu/mL), incubated in two different temperatures (35-37 °C and 22-24 °C) and different incubation times (18-96 h) evaluated by FCM. This three different dilutions were also inoculated into special platelet culture bottles (BacT/ALERT® BPA) and loaded into the blood culture system. Additionally 80 APSs routinely prepared in the Transfusion Center were evaluated by both FCM and the blood culture system. Platelets were lysed by freeze-thaw method.All spiked samples were positive with BacT/ALERT® BPA in 12-18 h. In 96 h incubation at 22-24 °C, the presence of bacteria was detected by FCM in all other samples (31/33) except low dilutions (1-10 and 10-100 CFU/ml) of K.pneumoniae standard strain. In the 35-37 °C, the presence of bacteria was detected by FCM in all samples (33/33) after 48 h of incubation. In routine APS one sample detected as positive (Bacillus simplex) with BacT/ALERT® BPA and no positivity was detected by FCM.The freeze-thaw method, which we have optimized for the lysis of platelets, is very practical and can be easily applied. The BacT/ALERT® system has been found to be very sensitive in detecting bacterial contamination in PSs. Flow cytometry method has been found to be successful, fast, easy to use and low cost in detecting bacterial contamination in PSs.

An evaluation of the clinical impacts of 7-day platelets

BACKGROUND

In August 2017, Canadian Blood Services extended the shelf-life of platelet concentrates from 5 to 7 days. The clinical impacts of this policy change remain unclear.

STUDY DESIGN AND METHODS

We used a before-after retrospective design of platelet-transfused adult inpatients in Hamilton, ON, Canada. Data were captured for 18 months before (Period 1: February 2016-July 2017) and 18 months after (Period 2: September 2017-February 2019) 7-day platelet implementation. Primary outcome was absolute platelet count increment (ACI) in univariate and multivariate analyses adjusted for confounders. Data were obtained from our institution's transfusion database, Ontario's Transfusion Transmitted Injuries Surveillance System, and the blood supplier.

RESULTS

Overall, 1360 patients with single dose platelet transfusions were included in Period 1 and 1211 patients in Period 2. Median age at admission was 66 years, and approximately 40% of patients underwent cardiac surgery. Using a non-inferiority margin of -10 × 10⁹ /L, platelets transfused during the 7-day storage period were non-inferior to those transfused in the 5-day storage period [mean count difference - 4.63 × 10⁹ /L (95% CI -7.40 to -1.87, p = 0.0001)]. However, platelet ACIs following transfusion consistently trended lower in the 7-day group for all patients and subgroups. No differences in secondary clinical outcomes were observed. Platelet expiry reduced from 8.1 to 6.3% (p < 0.0001).

CONCLUSION

Platelet transfusions following 7-day storage policy were non-inferior to transfusions in the 5-day policy period, although reduced ACIs were observed. There were no increases in adverse clinical outcomes.

Current status of rapid bacterial detection methods for platelet components: A 20-year review by the ISBT Transfusion-Transmitted Infectious Diseases Working Party Subgroup on Bacteria

BACKGROUND AND OBJECTIVES

Bacterial contamination of platelet components (PCs) poses a safety challenge for transfusion patients. Despite mitigation interventions, the residual risk of transfusion-transmitted bacterial infections remains predominant. PC safety can be improved either by pathogen reduction or by implementation of bacterial detection methods. Detection methodologies include culture methods and rapid detection methods. The current review focuses on currently available rapid detection methods.

MATERIALS AND METHODS

We reviewed published manuscripts since 2000 on rapid bacterial detection methods used for PC screening with result determination within 4 h. Methods meeting this criterion included Verax PGDprime, BacTx and nucleic amplification testing. The analytical and diagnostic sensitivity and specificity of these systems were assessed.

RESULTS

The analytical sensitivity between the different detection methods ranged between 50 and 100,000 CFU/ml. The sample volume used by these testing systems varies between 0.5 and 1.0 ml of PCs. A delay of at least 48 h before sampling enhances detectability. All rapid detection methods generate results in a timely manner, allowing testing to be performed before transfusion with optimal sensitivity.

CONCLUSION

Rapid detection methods improve PC safety regarding bacterial contamination. The assays are optimal for rapidly growing bacteria, which are more likely to cause septic transfusion reactions in patients. Because of the reduced diagnostic sensitivity, the sample collection should be late in shelf-life and ideally just before transfusion. The major benefit of these methods is that the test result can be obtained before releasing PCs for transfusion or to be used in combination with other screening methods applied early during PC storage.

Sterility Testing of Platelets Concentrate within Quality Control: Experiences and Opportunities to Extend the Application

Despite numerous measures, bacterial sepsis associated with the transfusion remains a major threat. The incidence of septic events induced by platelets transfusion is approximately 10 times higher than with transfused red blood cells due to their storage temperature. This caused new Standard that implements the methods for the detection and reduction of bacteria in the platelet concentrates (PC). The aim is to consider the possibility of wider application of this tests in order to extend the shelf-life of PC. Sterility testing of PC is done once or twice per month using BacT/Alert BPA and BacT/Alert BPN bottles. If positive, all products from the initial unit were tested to confirm or deny the status. During six years period, 67236 PC units were made and 872 of them were tested. Only two were found initially positive. After testing the other products from the same initial unit, results were negative so, final results proclaimed false positive. Pretransfusion bacterial detection is an important potential method for reducing the risk of bacteriemia and transfusion-associated septic reactions. In addition to routine measures, Mirasol PRT pathogen inactivation system, could be included. This allows certain amount of PC to be inactivated during the first 32 hours. Untreated PC units would be stored in standard conditions and for given time (three days) potentially present bacteria would reach a detectable level. This way the quantity of samples for sterility testing could be reduced, taking only 2 mL of each of four units of PC. Samples would be planted at the same vial-aerobic bottle, which would also, double the capacity in BacT/Alert 3D automated system.

A dual-center evaluation of platelet culture vials to detect the presence of microorganisms in platelets

BACKGROUND

Microorganism contamination of platelets results in a high risk of transfusion-related sepsis. Here, the ability of culture vials (BD BACTEC Platelet Aerobic/F and Platelet Anaerobic/F vials, Becton, Dickinson and Company) to detect microorganisms in leukoreduced apheresis platelets (LRAPs) and leukoreduced whole blood platelet concentrates (LRWBPCs) was assessed.

METHODS

LRAPs or LRWBPCs were inoculated into Aerobic/F and Anaerobic/F vials and placed in a blood culturing system (BD BACTEC FX System, Becton, Dickinson and Company) for growth/monitoring over 7 days to detect preexisting contamination during false-positive testing. Subsequently, platelets were seeded with microorganisms at approximately 10 CFU/mL or approximately 1 CFU/mL to simulate contamination. Aerobic/F and Anaerobic/F vials were inoculated with platelets (sets of 12). Microorganism growth was detected in the BACTEC FX instrument over 7 days. Overall, 2925 vials were tested.

RESULTS

Of the 1905 vials included in the microorganism detection phase, 63 (3.3%) Aerobic/F and 16 (0.8%) Anaerobic/F vials were both BACTEC FX and subculture negative. From the remaining 1827 vials, two (0.1%) Anaerobic/F vials were false positive; no false positives were observed in Aerobic/F vials, and no false negatives occurred in either vial type. Of the remaining 1825 vials (99.9%), 955 Aerobic/F and 870 Anaerobic/F vials were true positives. The mean-time-to-detection range was 8.5 to 77 hours. All true-positive Aerobic/F and Anaerobic/F vials showed 100% agreement with subculture for positive identification of seeded microorganisms.

CONCLUSION

Aerobic/F and Anaerobic/F vials facilitate contamination detection in LRAPs and LRWBPCs down to approximately 1 CFU/mL. These results support the use of Aerobic/F and Anaerobic/F vials for quality control testing of platelets before transfusion.

Detection of bacterial contamination in platelet concentrates from Brazilian donors by molecular amplification of the ribosomal 16S gene

OBJECTIVE

The aim of our work was to establish a semi-automated high-throughput DNA amplification method for the universal screening of bacteria in platelet concentrates (PCs).

BACKGROUND

Among cases of transfusion transmission of infectious agents, bacterial contamination ranks first in the number of events, morbidity and mortality. Transmission occurs mainly by transfused PCs. Automated culture is adopted by some blood banks for screening of bacterial contamination, but this procedure is expensive and has a relatively long turnaround time.

METHODS

PCs were spiked with suspensions of five different bacterial species in a final concentration of 1 and 10 colony-forming units (CFU) per millilitre. After incubation, the presence of bacteria was investigated by real-time polymerase chain reaction (PCR) and by the Enhanced Bacterial Detection System (eBDS, Pall) assay as a reference method. Real-time PCR amplification was performed with a set of universal primers and probes targeting the 16S rRNA gene. Co-amplification of human mitochondrial DNA served as an internal control.

RESULTS

Using the real-time PCR method, it was possible to detect the presence of all bacterial species tested with an initial concentration of 10 CFU mL^-1 24 h after contamination, except for Staphylococcus hominis. The PCR assay also detected, at 24 h, the presence of Serratia marcescens and Enterobacter cloacae with an initial concentration of 1 CFU mL^-1 .

CONCLUSIONS

The real-time PCR assay may be a reliable alternative to conventional culture methods in the screening of bacterial contamination of PCs, enabling bacterial detection even with a low initial concentration of microorganisms.

Pediatric red cell and platelet transfusions

The aim of pediatric transfusions should be based on the concept of avoiding unnecessary transfusions without jeopardizing the patient safety and providing correct blood components when there are well founded indications to transfuse. Despite considerable efforts from transfusion services to increase transfusion safety, transfusions are still associated with preventable and unpreventable adverse effects that may, in the worst case, have severe and fatal consequences. Transfusions to pediatric patients constitute a small proportion of all transfusions but have higher incidence of adverse events compared to adults. Pediatric transfusions consist of intrauterine transfusions, top-up transfusions to neonates and young children, exchange transfusions in the management of hemolytic disease of newborn (HDN), in addition to sickle cell crisis, chronic transfusion therapy in thalassemia patients, massive transfusion in trauma, HLA- and HPA-compatible platelets in immunized patients and neonates with fetal neonatal alloimmune thrombocytopenia (FNAIT). Packed red cells (PRCs) and platelet (PLT) concentrates are the most utilized blood components and will be reviewed here.

Acid sphingomyelinase mediates murine acute lung injury following transfusion of aged platelets

Pulmonary complications from stored blood products are the leading cause of mortality related to transfusion. Transfusion-related acute lung injury is mediated by antibodies or bioactive mediators, yet underlying mechanisms are incompletely understood. Sphingolipids such as ceramide regulate lung injury, and their composition changes as a function of time in stored blood. Here, we tested the hypothesis that aged platelets may induce lung injury via a sphingolipid-mediated mechanism. To assess this hypothesis, a two-hit mouse model was devised. Recipient mice were treated with 2 mg/kg intraperitoneal lipopolysaccharide (priming) 2 h before transfusion of 10 ml/kg stored (1–5 days) platelets treated with or without addition of acid sphingomyelinase inhibitor ARC39 or platelets from acid sphingomyelinase-deficient mice, which both reduce ceramide formation. Transfused mice were examined for signs of pulmonary neutrophil accumulation, endothelial barrier dysfunction, and histological evidence of lung injury. Sphingolipid profiles in stored platelets were analyzed by mass spectrophotometry. Transfusion of aged platelets into primed mice induced characteristic features of lung injury, which increased in severity as a function of storage time. Ceramide accumulated in platelets during storage, but this was attenuated by ARC39 or in acid sphingomyelinase-deficient platelets. Compared with wild-type platelets, transfusion of ARC39-treated or acid sphingomyelinase-deficient aged platelets alleviated lung injury. Aged platelets elicit lung injury in primed recipient mice, which can be alleviated by pharmacological inhibition or genetic deletion of acid sphingomyelinase. Interventions targeting sphingolipid formation represent a promising strategy to increase the safety and longevity of stored blood products.

Bacterial screening of platelet components by National Health Service Blood and Transplant, an effective risk reduction measure

BACKGROUND

Bacterial contamination of blood components remains a major cause of sepsis in transfusion medicine. Between 2006 and 2010 in the 5 years before the introduction of bacterial screening of platelet (PLT) components by National Health Service Blood and Transplant (NHSBT), seven cases of PLT component-associated transmission of bacterial infection were recorded for 10 patients, three of which were fatal.

STUDY DESIGN AND METHODS

Sampling of individual PLT components was undertaken at 36 to 48 hours after donation and tested in the BacT/ALERT system with 8 mL inoculated into each of aerobic and anaerobic culture bottles. Bottles were incubated until the end of the 7-day shelf life and initial reactive bottles were examined for contamination. Bacterial screened time-expired PLTs were tested as in the screen method.

RESULTS

From February 2011 to September 2015, a total of 1,239,029 PLT components were screened. Initial-reactive, confirmed-positive, and false-positive rates were 0.37, 0.03, and 0.19%, respectively. False-negative cultures, all with Staphylococcus aureus, occurred on four occasions; three were visually detected before transfusion and one confirmed transmission resulted in patient morbidity. The NHSBT screening protocol effectively reduced the number of clinically adverse transfusion transmissions by 90% in this reporting period, compared to a similar time period before implementation. Delayed testing of 4515 time-expired PLT units after screening revealed no positives.

CONCLUSION

The implementation of bacterial screening of PLT components with the NHSBT BacT/ALERT protocol was an effective risk reduction measure and increased the safety of the blood supply.

Storage medium of platelet transfusions and the risk of transfusion-transmitted bacterial infections

BACKGROUND

Transfusion-transmitted bacterial infections (TTBIs) are among the most concerning risks of transfusion of platelet (PLT) concentrates. Storage medium influences bacterial growth dynamics and thereby the sensitivity of screening tests for bacterial contamination.

STUDY DESIGN AND METHODS

The aim of this study was to quantify the association of storage media with the incidence of TTBIs after transfusion of PLT concentrates. In the Netherlands, the choice of storage medium is determined solely by geographic location of the hospital. We compared types of storage medium of all reported cases of TTBIs after transfusion of a PLT concentrate with types of storage medium of all produced PLT concentrates in the Netherlands from 2003 to 2014.

RESULTS

Fourteen cases of TTBIs were reported, of which 57.1% received a PLT concentrate stored in PLT additive solution (PAS) and 42.9% a PLT concentrate stored in plasma. Of all produced PLT concentrates 22.3% were stored in PAS and 77.7% in plasma. The relative risk of TTBI after transfusion of a PAS-stored PLT concentrate was 4.63 (95% confidence interval [CI], 1.4-16.2) compared to transfusion of a plasma-stored PLT concentrate. The incidence of TTBIs was 22.2 per million (95% CI, 12.1-37.2 per million) transfused buffy coat PLT concentrates.

CONCLUSION

Transfusion of PAS-stored PLT concentrates is associated with a fourfold increased incidence of TTBIs, compared to plasma-stored PLT concentrates.

Platelet Septic Transfusion Reactions in Patients With Hemato-Oncological Diseases

BACKGROUND & OBJECTIVE

Bacterial, contamination of blood components are a significant risk for transfusion reactions. Inherently, platelet concentrates (PCs) are vulnerable to bacterial contamination, due to the storage condition of processed PCs at room temperature, which provide very suitable conditions for the proliferation of microorganisms.The current study aimed at investigating the transfusion associated septic reaction rate in patients with hemato-oncological diseases in Imam Khomeini Hospital, Tehran, Iran, and identifying the contaminating bacteria.

METHODS

A total of 3056 adult patients of the Cancer Center of Imam Khomeini Hospital in Tehran transfused with PCs were studied based on the clinical symptoms of septic transfusion reaction from June 1, 2010 to May 31, 2011. Patient presented with the criteria of reaction and the residual components were evaluated for bacterial contamination by Bac T/Alert system.

RESULTS

Patients with leukemia or lymphoma transfused with random-donor PCs were evaluated the signs and symptoms of transfusion reaction occurred only in 12 (%0.4) cases. Automated cultivation found 3 positive blood cultures. Among these a male recipient was categorized as possible septic transfusion reaction and Citrobacterfreundii was isolated from blood sample.

CONCLUSION

Appropriate clinical utilization of PCs transfusion, and ongoing vigilance to recognize, investigate, promptly treat, and report all suspicious transfusion reactions are necessary to manage the transfusion complication including transfusion-transmitted infections (TTI).

Soluble Mediators in Platelet Concentrates Modulate Dendritic Cell Inflammatory Responses in an Experimental Model of Transfusion

The transfusion of platelet concentrates (PCs) is widely used to treat thrombocytopenia and severe trauma. Ex vivo storage of PCs is associated with a storage lesion characterized by partial platelet activation and the release of soluble mediators, such as soluble CD40 ligand (sCD40L), RANTES, and interleukin (IL)-8. An in vitro whole blood culture transfusion model was employed to assess whether mediators present in PC supernatants (PC-SNs) modulated dendritic cell (DC)-specific inflammatory responses (intracellular staining) and the overall inflammatory response (cytometric bead array). Lipopolysaccharide (LPS) was included in parallel cultures to model the impact of PC-SNs on cell responses following toll-like receptor-mediated pathogen recognition. The impact of both the PC dose (10%, 25%) and ex vivo storage period was investigated [day 2 (D2), day 5 (D5), day 7 (D7)]. PC-SNs alone had minimal impact on DC-specific inflammatory responses and the overall inflammatory response. However, in the presence of LPS, exposure to PC-SNs resulted in a significant dose-associated suppression of the production of DC IL-12, IL-6, IL-1α, tumor necrosis factor-α (TNF-α), and macrophage inflammatory protein (MIP)-1β and storage-associated suppression of the production of DC IL-10, TNF-α, and IL-8. For the overall inflammatory response, IL-6, TNF-α, MIP-1α, MIP-1β, and inflammatory protein (IP)-10 were significantly suppressed and IL-8, IL-10, and IL-1β significantly increased following exposure to PC-SNs in the presence of LPS. These data suggest that soluble mediators present in PCs significantly suppress DC function and modulate the overall inflammatory response, particularly in the presence of an infectious stimulus. Given the central role of DCs in the initiation and regulation of the immune response, these results suggest that modulation of the DC inflammatory profile is a probable mechanism contributing to transfusion-related complications.

Diagnostic methods for platelet bacteria screening: current status and developments

Bacterial contamination of blood components and the prevention of transfusion-associated bacterial infection still remains a major challenge in transfusion medicine. Over the past few decades, a significant reduction in the transmission of viral infections has been achieved due to the introduction of mandatory virus screening. Platelet concentrates (PCs) represent one of the highest risks for bacterial infection. This is due to the required storage conditions for PCs in gas-permeable containers at room temperature with constant agitation, which support bacterial proliferation from low contamination levels to high titers. In contrast to virus screening, since 1997 in Germany bacterial testing of PCs is only performed as a routine quality control or, since 2008, to prolong the shelf life to 5 days. In general, bacterial screening of PCs by cultivation methods is implemented by the various blood services. Although these culturing systems will remain the gold standard, the significance of rapid methods for screening for bacterial contamination has increased over the last few years. These new methods provide powerful tools for increasing the bacterial safety of blood components. This article summarizes the course of policies and provisions introduced to increase bacterial safety of blood components in Germany. Furthermore, we give an overview of the different diagnostic methods for bacterial screening of PCs and their current applicability in routine screening processes.

Bacterial safety of cell-based therapeutic preparations, focusing on haematopoietic progenitor cells

Bacterial safety of cellular preparations, especially haematopoietic progenitor cells (HPCs), as well as advanced therapy medicinal products (ATMPs) derived from stem cells of various origins, present a challenge for physicians, manufacturers and regulators. The article describes the background and practical issues in this area and illustrates why sterility of these products cannot currently be guaranteed. Advantages and limitations of approaches both for classical sterility testing and for microbiological control using automated culture systems are discussed. The review considers novel approaches for growth-based rapid microbiological control with high sensitivity and faster availability of results, as well as new methods for rapid bacterial detection in cellular preparations enabling meaningful information about product contamination within one to two hours. Generally, however, these direct rapid methods are less sensitive and have greater sampling error compared with the growth-based methods. Opportunities for pyrogen testing of cell therapeutics are also discussed. There is an urgent need for development of novel principles and methods applicable to bacterial safety of cellular therapeutics. We also need a major shift in approach from the traditional view of sterility evaluation (identify anything and everything) to a new thinking about how to find what is clinically relevant within the time frame available for the special clinical circumstances in which these products are used. The review concludes with recommendations for optimization of microbiological control of cellular preparations, focusing on HPCs.

[Prevention of bacterial risk: pathogen inactivation/detection of bacteria]

Bacterial contamination of blood products remains the most important infectious risk of blood transfusion in 2013. Platelet concentrates (PC) are in cause in the majority of the transfusion reaction due to bacterial contaminations. A lot of prevention methods have been developed over the last 10 years (pre-donation interview, skin decontamination, diversion of the first 30 mL of the donation, leuko-reduction...), they have focused on limiting the contamination of the donations and prevent the bacterial growth in donations and/or in the blood products. These measures were effective and led to significantly reducing the risk of adverse effects associated with bacterial growth. However, every year there are about six accidents (with a high level of imputability) and one death. The reduction of the bacterial risk remains a priority for the French Blood Establishment (EFS). The procedure for skin disinfection is going to be improved in order to further strengthen this crucial step to avoid the contamination of donation. Methods of pathogen inactivation applied to plasma and PC are available in France and their effectiveness is demonstrated on the bacterial risk. Methods for bacterial detection of PC are used in many countries now. Automated culture is the most common. Alternatives are now available in the form of rapid tests able to analyze the PC just before the delivery and avoid false negatives observed with automated culture. Assessments are under way to confirm these benefits in 2013.

[Control of the bacterial risk of transfusion in France in 2013]

Bacterial contamination of blood products (BP) remains the most important infectious risks of blood transfusion in 2013. Platelet concentrates (PC) are the blood products the most at risk, whether CPA or MCPS. In France, the residual risk has been steadily declining since 1994. For the platelets, the frequency of transfusion reaction due to bacterial contamination (TRBC) is now about at one per 50,000 CP distributed. The number of deaths has remained stable since 1994 with one death per year (300,000 distributed CP). The progressive decrease in the number of cases of TRBCs is the result of steady improvement of practices and prevention methods at all stages from collection to the transfusion of BP. But if all these improvements have significantly reduced the incidence of TRBCs, mortality is not changed with the CP and the reduction of this risk is a priority for the French Blood Establishment (EFS). Detection methods of CP contaminated or pathogen inactivation are two approaches available and can provide a significant reduction (for the former) or deletion (for seconds) of the risk of transfused contaminated CP. Currently, the choice is in favor of the detection of bacteria. New detection "rapid tests" methods were added to the panel of candidates and are being evaluated. Inactivation of pathogens remains the safest prospect of eliminating this adverse effect of transfusion. Implementation of one method for bacterial detection is probably a transitional measure.

Effects of platelet concentrate storage time reduction in patients after blood stem cell transplantation

OBJECTIVE

To evaluate the clinical effect of platelet concentrate (PC) transfusions after PC storage time reduction to 4 days.

PATIENTS AND METHODS

This was a single-centre cohort study comparing two 3-month periods of time, before and after the reduction of PC storage time from 5 to 4 days. Seventy-seven consecutive patients with PC transfusions were enrolled after blood stem cell transplantation. Corrected platelet count increment (CCI) on the morning after transfusion, time to next platelet transfusion, need for red blood cell (RBC) transfusion and clinical bleeding symptoms were compared.

RESULTS

Platelet concentrate storage time was reduced between period 1 (storage for up to 5 days, median storage time 78 h, range 11-136 h) and period 2 (storage for up to 4 days, median storage time 53 h, range 11-112 h). Patients were comparable for age, weight, body surface area, underlying disorder, type of transplantation and transfused platelet dose. The CCI increased from a median of 4 (range 0-20) to 8 (0-68) × 10(9) /l per 10(11) platelets/m(2) (P < 0·0001). Time to next PC transfusion increased from 1·1 to 2·0 days (P < 0·0001). Any bleeding symptom was noted in 20 of 36 patients (56%) vs. 9/41 patients (22%, P < 0·01). Nose bleeds, haematuria and bleeding at more than one site were significantly reduced. Frequency of RBC transfusion within 5 days after PC transfusion was reduced from 74 to 58% (P < 0·0001).

CONCLUSION

Platelet concentrate storage time shortening was associated with highly significant CCI increase, reduced RC needs and lower patient numbers with bleeding events.