[Analysis of causes of destruction of labile blood products in health institutions: A multicentric study]. / Analyse des causes de destruction des produits sanguins labiles dans les établissements de santé : une étude multicentrique.

OBJECTIVE: One of the tasks of haemovigilance correspondents in Health Institutions (HI) is to reduce the destruction of labile blood components (LBC). The objective of this study was to analyse in depth, five years after a first multicentric study, the causes of LBC destruction in order to assess the impact of measures taken and to define new ways of improvement. METHODS: Prospective analysis was carried out throughout 2016. For every LBC destroyed, the following elements were reported: type of LBC, transfusion department, cause of destruction analysed according to a decision tree, subsequently classed as avoidable or unavoidable. RESULTS: The study included 15 HI. A total 3058 LBC were destroyed, representing an average 0.90% of issued LBC, and this analysis concerned 2576 LBC. Sixty-seven percent of LBC were issued for surgery, intensive care or emergencies. Forty percent of the causes of destruction were patient-related (death, clinical worsening, adverse effects or abnormal constants prior to delivery). Thirty percent were prescription-related, mainly cases of excessive prescription for different reasons. Eleven percent were linked to organisational issues. The rate of destruction judged avoidable, all causes combined, was 36%. CONCLUSION: Comparison with the precedent study shows improvement, thus revealing the efficacy of implemented measures (single-dose distribution, return procedures back to the site of distribution, training of participants). In order to further reduce this rate of destruction, we suggest to promote storage procedures and, above all, to continue to raise awareness within healthcare teams.

[Is the research of posttransfusional alloantibodies still relevant?] / Est-ce que la réalisation de la recherche d'anticorps antiérythrocytaires post­transfusionnelle est toujours pertinente ?

The decision of November 6th, 2006 defining the principles of best practices recommends that posttransfusional red cell alloantibodies research is performed after one to three months after. In the University hospital of Brest, the haemovigilance unit takes charge of sending the medical prescription within the required time and centralizing the results. We wished to estimate if the realization of this research still remains relevant. METHODS: A prospective analysis was performed in 2015. We evaluated the realization rate, the red cell alloantibodies rate and the recipient adverse reactions with the diagnostic category: alloimmunization (delayed serological transfusion reaction, DSTR). RESULTS: In 2015, 2162 prescriptions were sent to the 3271 transfused patients. One thousand and eighteen red cell alloantibodies research were done, i.e. a return rate of 61%. Among them, 12 alloantibodies appeared (0.9%) within an average of 56 days. Thirty-three other alloantibodies appeared and were discovered most frequently before a new transfusion. In 10 cases, a posttransfusional research was done that was negative. A survey was conducted among GHCOH members to describe the practices in these health institutions. Twelve questionnaires were analysed. Ten institutions performed a posttransfusional alloantibodies research by issuing a prescription at the patient's exit with a return rate between 0.14 and 16%; 1 institution has a centralized organization with a return rate of 68.3%; 1566 red cell alloantibodies research were performed and among them, 24 alloantibodies appeared (1.53%). CONCLUSION: These results indicate that to be effective, the management of this biological test must be centralized. Despite this, the red cell alloantibodies rate remains very low (0.9 and 1.53%) and raises the question of the relevance of this systematic testing after transfusion, which is in any case mandatory before a new transfusion of red blood cells.

[French haemovigilance from 1994 to nowadays: Evolution and prospects]. / L'hémovigilance française de 1994 à nos jours : évolution et perspectives.

Born in France in 1993, haemovigilance aims at monitoring the blood transfusion chain from the donor to the recipient, receiving labile blood products. It differs from the other vigilances due to its organization and its completeness dealing with the collected information. Prior to the European directive, the French agency created a new gravity level: grade 0, aiming at reporting and analyzing the blood transfusion chain dysfunction. Thus, haemovigilance gradually integrates the management of the risks and is given new missions in hospitals and private hospitals. One of its first actions will be to achieve an a-priori analysis, preventing the risks throughout the blood transfusion chain. Such crosscutting missions will be used to manage the coordination of vigilance and in some cases identity monitoring. Haemovigilance may be a key player in identity monitoring because patients' immunology-hematological data base can make it possible to confirm or deny a patient's identity (misused identity or homonyms). Haemovigilance is going to meet other challenges such as the training of health professionals', the implementation of patients' blood management and the periodic revision of the blood transfusions bulletin. A new crosscutting medical profession appears in some countries: blood transfusion practitioner. It combines vigilance, risk management, support for the therapeutic blood transfusions, health professionals training and the evaluation of practices and results. A final mission would be for haemovigilance to be responsible for medicinal products derived from human blood to allow a better monitoring of plasma transfusions.

Evaluation of the non-compliance with grouping guidelines which may lead to "wrong blood in tube", an observational study and risk factor analysis.

BACKGROUND: In France, blood group determination requires the completion of two samples collected at two different times to detect identity mistake and "wrong blood in tube". The aims of the present study were: (1) to evaluate the compliance with guidelines and (2) to identify risk factors of non-compliance. MATERIALS AND METHODS: Samples for ABO group determination collected between January 1st and December 15th, 2013 in the University hospital of Nîmes, France were analyzed. An ABO group determination demand was considered non-compliant if more than one tube arrived in the laboratory within ten minutes apart. Between May 1st and June 30th 2014, a self-administered questionnaire was offered to the nurses of the hospital on a random day for each service during this period. The aim was to validate the non-compliance criterion and the identification of risk factors using logistic regression. RESULTS: Among the 16,450 analyzed blood samples, the overall compliance rate was 65.1%. Lower compliance rates were found in the surgical services. Independent risk factors for wrong practice were work overload, surgical service and individual intermediate transfusion frequency. DISCUSSION: More than one third of ABO group determinations did not follow national recommendations, which induces a substantial risk of "wrong blood in tube" and group error. The study revealed major variations among hospital services. Identification of risk factors allows targeted corrective actions.

[Computer assisted prescription of labile blood products: What are we expecting?] / L'apport de la prescription assistée par ordinateur des produits sanguins labiles.

Computer assisted prescription of labile blood products is just at its beginning. Current programs already allow embedding automatically such data as patient's and prescribers' identification or ward details to produce readable prescriptions, also complying with part of Good Practice guidelines. Now prescriptions can also be sent electronically to the Etablissement Francais du Sang, the French blood products services. Usually they are computer programs specialised in transfusion and interfaced with the main patient's file software. Hardly ever the main software is able to manage transfusion itself. Next step would consist in performing checks, calculations or displaying warning or help messages based on academic or local medical recommendations or even tailored to pre-defined individual requirements. But these call for direct access to patient's data such as diagnosis or tests results, that must be accurately classified and coded before use. The main software could provide such functionalities: but actually that would be infrequent and difficult to transpose from one hospital to the other, regarding to the diversity of main software and their settings. Another solution would be to enhance the very few transfusion specialised programs in order to assist prescribers. Data could be prepared and sent by the main software according to a standardised format each time a prescription is to be entered. This standardised format should be independent from software in order to ensure interoperability, whatever the main and specialised programs. The content and format of this data exchange has to be defined, but this would allow hundreds of hospitals to provide a comprehensive tool for prescription of labile blood products, regardless of their main patient's file software.

[Blood products delivery by hospital blood banks in emergency situations in France]. / Délivrance en urgence des produits sanguins labiles à partir des dépôts de sang.

In France, 13.5% of labile blood products (LBP) are delivered by 644 hospital blood banks, 472 being devoted to emergency only. A nation-wide survey aimed to describe their activity in emergency situations was carried out from September 12th to 26th, 2011. All of 26 French regions but two participated, 548 (85%) of their hospital blood banks answered a questionnaire, of which 56% had actually issued blood in emergency situations. A total of 5241 LBP were delivered, 24% for utmost vital emergencies (UVE), 25% for vital emergencies (VE) and 51% for relative emergencies (RE). In UVE and VE, 47% of the packed red blood cells (PRBC) delivered were O RH1. Females were half of the 2102 patients, mean age was 67 years. In UVE, banks devoted to emergency only, issued LBP for 228 patients (57%) and other banks for 169, acute digestive tract and surgical hemorrhage being the most frequent indications. The average number of PRBC transfused in 24 hours per patient varied with the criticality of the emergency: 2.3 for RE, 3.9 for VE and 6.1 for UVE. The mortality rate at 24 hours of first delivery was 1.2% in RE, 2.8% in VE and 10.1% in UVE. Time between prescription and start of transfusion of PRBC in UVE was within 15 minutes in 50% of cases and within 25 minutes in 75% of cases for emergency devoted banks but respectively 25% and less than 50% for others. The proportion of surviving patients having received more than three LBP in 24 hours was 58% in UVE, 33% in VE and 10% in RE. The proportion of over the limit waiting time for PRBC transfusion in UVE is critical, mainly for banks not only devoted to emergencies. The high proportion of UVE and VE patients receiving three PRBC or less in 24 hours may also be an indicator that some physicians do not comply with emergency prescription rules. An effort should be made to improve the activity, in emergencies, of a part of hospital blood banks and the way physicians use them.

[Data processing and blood transfusion activities: fact and future in 2013]. / La place de l'informatique dans l'activité transfusionnelle et son évolution--Panorama 2013.

It is now hard to think of blood transfusion activities without data processing. Blood transfusion centers are unable to work without it since a long time. Its necessity in hospital blood banks is following the same pattern. Electronic data interchange between them is growing because of their high interdependence. A lot has already been done and works routinely. But a lot remains to be done, due to continuous evolution of computer science and blood transfusion itself.

[Computerization of hospital blood banks in France]. / Informatisation des dépôts de produits sanguins labiles en France.

In France, most blood products are delivered by the établissement francais du sang, directly to the recipients, and hospital blood banks deliver a minor part, but are independent from it. However that may be, hospital blood banks are hazardous activities regarding to recipients, blood products, blood supply of the hospital and regional blood supply. Because of the high risk level, a computerized information system is compulsory for all hospital blood banks, except for those only devoted to vital emergency transfusion. On the field, the integration of computerization in the different processes is very heterogeneous. So, it has been decided to publish guidelines for computerizing hospital blood banks information systems and production management. They have been built according to risk assessment and are intended to minimize those risks. The principle is that all acquisition and processing of data about recipients or blood products and tracking, must be fully computerized and that the result of all manual processes must be checked by computer before proceeding to the next step. The guidelines list the different processes and, for each of them, the functions the software must play. All together, they form the basic level all hospital blood banks should reach. Optional functions are listed. Moreover, the guidelines are also aimed to be a common tool for regional health authorities who supervise hospital blood banks.

[Computerization of hospital blood banks in France: what has been done and what remains to be done]. / Informatisation des dépôts de sang: ce qui est fait et ce qui reste à faire.

French regulation requires hospital blood banks to have data processing of their blood transfusion activities. In order to help them to achieve this goal, guidelines were published by an official working party. For their diffusion, a study was launched to assess current gaps between guidelines and reality. Analysis of the first forty responses shows: 1) computerization is already set up for main blood banks and is going well for the others; 2) most of guideline functionalities are now working in main blood banks; but some points need specific developments; 3) taken as a whole, other blood banks are further from guideline requirements, even if some have already reached the goal; 4) guidelines are able to point out to all hospital blood banks what remains to be done.

[Specific hazards of transfusion in medical units]. / Risques spécifiques de la transfusion en médecine.

In France, more than 1100 hospitals have medical wards with specialty units that perform transfusions and use about 40% of all labile blood products. Due to patients suffering from multiple pathologies, repeated transfusions or use of platelet concentrates, these hospital units encounter specific and more frequent transfusion hazards than other wards. The incidence rate of transfusion associated cardiac overload, anti-erythrocyte, -HLA or -HPA allo-immunisations or transfusion-induced hemochromatosis, is high. Platelet concentrates may also cause severe allergic reactions. They are still under-reported, though immediate or delayed effects on patients or care processes are considerable. The prevention of these adverse reactions or of their consequences requires careful prescription, properly planned and organised patient stays, particularly at the day hospital, with special attention to coordination of doctors and nurses. Labelling errors of blood group samples or blood administration to the wrong patient are the most frequent serious adverse events. Main underlying causes are: a poor understanding of specific hazards of transfusion in medical wards, routine, multiple transfusions and permanent or occasional working overload. So, both doctors and nurses should do their best to improve organisation and systematically comply with good clinical practice.

[Factsheets: support in the analysis of recipients' adverse reactions]. / Fiches techniques: aide à l'analyse des effets indésirables receveurs.

In order to help the analysis of adverse effects of transfusion, factsheets have been written by working groups of the French agency for the safety of health products ANSM. Each factsheet deals with a blood transfusion side effect and is composed of five parts, including pathophysiological mechanisms, diagnostic criteria, management recommendations, etiologic investigations and rules for filing the notification form to ANSM. Since 2006, 11 factsheets have been published on the French haemovigilance network website. The major characteristics of the two last sheets published "post-transfusion purpura" and "non erythrocyte incompatibility reaction" are presented, followed by the updated card for "allergy". These factsheets give relevant guidelines allowing better evaluation of recipients' adverse reactions, particularly their diagnosis, severity and accountability. They also could initiate studies among European and international haemovigilance networks.

[Computerization of hospital blood banks in France]. / Informatisation des dépôts de produits sanguins labiles en France.

In France, most blood products are delivered by the Établissement français du sang, directly to the recipients, and hospital blood banks deliver a minor part, but are independent from it. However that may be, hospital blood banks are hazardous activities regarding recipients, blood products, blood supply for the hospital and regional blood supply. Because of the high risk level, a computerized information system is compulsory for all hospital blood banks, except for those only devoted to vital emergency transfusion. On the field, integration of computerization in the different processes is very heterogeneous. So it has been decided to publish guidelines for computerizing hospital blood banks information systems and production management. They have been built according to risk assessment and are intended to minimize those risks. The principle is that all acquisition and processing of data about recipients or blood products and tracking, must be fully computerized and that the result of all manual processes must be checked by computer before proceeding to the next step. The guidelines list the different processes and, for each of them, the functions the software must play. All together, they form the basic level all hospital blood banks should reach. Optional functions are listed. Moreover, the guidelines are also aimed at being a common tool for regional health authorities who supervise hospital blood banks.

[Reporting and notification of transfusion serious adverse events in France]. / Le signalement et la déclaration des incidents graves de la chaîne transfusionnelle.

Surveillance of serious adverse events in transfusion has recently been enforced in all European Union member states, and in France in 2007. This is an important expansion of the field of haemovigilance towards risk management in transfusion from donor selection to recipients follow up. Participation in this notification system is mandatory for all health professionals. They have to report immediately all serious adverse events to the haemovigilance officer of their hospital or transfusion centre. In accordance with a national procedure, the latter has to notify all serious adverse events to regional and national competent authorities using a web-based application. They also have to perform a root case analysis and propose corrective measures. However, as the definition of the seriousness of adverse events is still imprecise, there is a dramatic heterogeneity in the notification level among regions. Therefore, basic criteria of evaluation of surveillance systems, such as sensitivity or representativeness, are not fulfilled. Next step will be the introduction of a more precise definition of seriousness, through a list of events to be notified, in order to achieve a quality level similar to that of transfusion adverse reactions surveillance.

[Yes, we should keep ABO agglutination test within bedside transfusion checks]. / Oui, il faut conserver le test d'agglutination ABO dans le contrôle ultime au lit du malade.

ABO incompatible transfusions are still a frequent cause of serious adverse transfusion reactions. Bedside check is intended to detect patient errors and prevent ABO mismatch. France is one of the few countries that includes ABO agglutination test for red blood cells in bedside checks. Evaluation of this ABO agglutination test, performed with a special card, shows that, on the field, despite frequent users' mishandling, it can detect up to 93% of ABO incompatibilities. This is not enough to rely on this sole test for bedside checks. But, linking it with an another test, currently, checks that the right blood is given to the right patient, rises the sensitivity of the whole bedside procedure up to an estimated 99.65%, for detection of ABO incompatibilities. This linkage has been introduced in the French regulation in 2003. Since then, the incidence of ABO incompatible transfusions has decreased dramatically and faster than in any other country, so France has now, probably, the lowest rate of ABO incompatible transfusions. The investigation of the few ABO accidents that still occur, shows that professionals have always bypassed this linkage. On the other hand, introducing bedside recipient and blood products barcode or radio-chip checks in all the 1500 French hospitals, though technically possible, would provide very little enhancement and lead to major difficulties and expenses. Linkage of ABO agglutination test to patient and blood checks within the bedside procedure has proved to be efficient and should be kept.

[The decrease of preoperative autologous transfusion in France has not been linked to an increase of homologous red cell concentrates]. / La diminution de l'autotransfusion programmée en France ne s'est pas accompagnée d'une augmentation de consommation de concentrés de globules rouges homologues.

OBJECTIVE: The use of preoperative autologous blood transfusion has dramatically decreased in France. The aim of this study was to evaluate the evolution of practice both of autologous and homologous Red Blood Cells (RBC) concentrates transfusion between 2002 and 2005, and to asses the consequences of the highlighted changes. MATERIAL AND METHODS: Data on blood transfusion are collected and validated nationwide by a network of regional coordinators of haemovigilance. For each hospital, from 2002 to 2005, the annual changes in the number of transfused homologous and autologous RBC have been evaluated. The ratio of preoperative autologous RBC, number of autologous RBC divided by the number of all RBC, has also been calculated. Hospitals have been split into three cohorts, according to their 2005 autologous RBC ratio. For each cohort, correlations between the variations of the number of autologous, homologous and total RBC in each hospital have been studied. RESULTS: Data have been validated for 22 French regions that performed 71.8% of the total French transfusion in 2004. In 2005, 1,831,544 labile blood products have been transfused in 1197 hospitals and clinics among which 379 have used preoperative autologous transfusion. A total of 37,289 autologous RBC have been transfused in 2003, 28,689 in 2004 (-23.1%) and 17,758 in 2005 (-38.1%). The first cohort of 269 hospitals had a ratio of autologous RBC under 3%, the second cohort of 38 hospitals, a ratio between 3 and 6%, while the third cohort of 72 hospitals had transfused 6% or more of autologous RBC. In the two first cohorts, non-surgical activities were so large that it was impossible to assess the changes in surgical use of transfusion. The third cohort, with a ratio of 6% or more, was essentially devoted to surgery (88% of beds). These hospitals and clinics have transfused 13,076 autologous RBC in 2002 and 8583 in 2005 (-34.4%). In this group, there was a statistical correlation between the decrease of autologous RBC and the decrease of total RBC (r(2)=0.36), and no increase in the transfusion of homologous RBC has been observed. During the same period, neither hospitals nor clinics showed any decrease of their surgical activity. The drop of autologous RBC transfusion led to a decrease of the total number of RBC transfused and thus, to a decrease of the global exposure to transfusion hazard. CONCLUSION: The present results confirmed a decline of preoperative autologous transfusion in France, between 2002 and 2005. Meanwhile no supplementary need of homologous RBC has been observed among hospitals, performing surgery that formerly had a high ratio of autologous RBC.

[Evaluation of the organization of pre and post transfusion screening of viral infection markers in blood recipients in France in 2001]. / Evaluation de l'organisation du dépistage sérologique pré- et post-transfusionnel en France en 2001.

In France, data collection related to blood recipient's viral infectious disease markers pre and post-transfusion is a legal requirement for hospitals. Our study aimed to evaluate the actual modalities of this extensive screening in 2001, six years after the Ministry of health issued recommendations. A questionnaire was sent to the haemovigilance correspondents in hospitals having transfused labile blood products (LBP) in 2001. A total of 1463 hospitals having transfused 85% of LBP in France responded. 82.4% of hospitals have written guidelines for pre-transfusion screening of viral markers, mainly for HIV and hepatitis C. A frozen repository storage is held by 23.9% of hospitals with storage durations between 1 to 40 years. 84% of hospitals have written guidelines for post-transfusion screening. The test prescriptions are mostly done by physicians from clinical services and they include in more than 80% of cases, HIV and HCV markers. Only 12% of hospitals recontact the patient in case of a no show. Even though 77.5% of responding hospitals have labile blood products recipients follow up processes, their effectiveness remains quite low, only 16% of recipients having test results available at the hospital.

[Evaluation of pre- and post-transfusion serological screening in France in 2001]. / Evaluation du dépistage sérologique pré- et post-transfusionnel en France en 2001.

Screening of labile blood products recipients for HIV and HCV has been performed in France since a government recommendation was issued in 1996. It has been designed to get transfusion related contamination of recipients through pre- and post-transfusion serological tests. Since then, residual risk has decreased dramatically and it was suspected that recommendations might sometimes be ignored. A nationwide survey has been done to measure the real screening rate and its cost efficacy ratio. In addition accuracy of tracability and recipients mortality has also been evaluated. A random sample of 1115 labile blood products among all the 1203 378 distributed during first semester of 2001 in France has been drawn. They have been matched with test results obtained in hospital files. Tracability has been considered accurate if name, surname and birth date of recipients were exactly the same both in hospital file and in the file of the Etablissement Français du Sang. A total of 1092 hospital files has been retrieved. Pre transfusion HIV and HCV tests have been performed in 58.5 % of cases, 95 % CI [55.6-61.5], and post-transfusion tests in 30.5 % [28.5-35.5] of cases. Only 19.5% [16.6-22.6] of recipients, not known to be dead 6 months after transfusion, have had both pre and post-transfusion tests. No HIV or HCV contamination has been notified by the Haemovigilance network during the same period. Accuracy rate of tracability was 96.25% [94.9-97.3]. Furthermore 35.8% [33-38.7] of recipients were found dead within 6 months after transfusion. A logistic regression analysis showed that the hospital area, the hospital size (more than 300 beds) and the annual amount of blood bags transfused in it (less than 5000) were factors independently associated with having a full pre and post-transfusion screening. Currently, the screening program may detect 0.14 cases of HIV and 0.05 HCV transfusion related contamination of recipient every year. The total cost of this program is about 20 million euro and the cost per case exceeds 110 million euro. The program will be of no use in case of an emerging transmitable disease. This program does not comply to any evaluation criteria of screening programs and its cost efficacy ratio is very poor.

[Validity of pretransfusion bedside agglutination tests for pretransfusional control]. / Etude régionale de validité, dans les services de soins, du test d'agglutination du contrôle ultime prétransfusionnel.

The incidence rate of ABO hazards of transfusion remains high in France. In this country, bedside pretransfusion controls include an agglutination test for red cells only, although its validity has scarcely been assessed in the practice. 847 nurses from 9 public hospitals and private clinics in a French region participated in a study aimed at measuring the sensitivity and specificity of pretransfusion bedside agglutination tests within hospital wards. Sensitivity was found to be 93.9% +/- 3%. Nondetection of mismatching was increased by two risk factors only: having worked more than 4 years in the same ward, and not having been trained to use this test. The sensitivity of this test might still be improved. This test is found sensitive enough to be kept. Nevertheless, if used alone, it is not a safe protection against recipient's mismatch. Authors recommend both to improve agglutination test sensitivity and to link it strongly to the bedside checking of both transfusion information and the recipient's identity.

[Epidemiological estimation of the effectiveness of the pertussis vaccination during outbreaks in a community. Les pédiatres et bactériologistes du réseau RENACOQ]. / Estimation épidémiologique de l'efficacité de la vaccination anticoquelucheuse au cours d'épidémies en collectivité. Les pédiatres et bactériologistes du réseau RENACOQ.

BACKGROUND: With the recent pertussis vaccine trials, the efficacy of acellular vaccines is now well known, estimated at 85% for multicomponent vaccines. On the other hand, the estimates of whole cell vaccines efficacy varies from 36% to 98% with the different vaccines used. We evaluated the field effectiveness of the French whole cell pertussis vaccine during outbreaks in schools and centers for disabled children. METHODS: Four limited outbreaks between 1993 and 1995 were investigated using a retrospective cohort study design. Vaccine effectiveness (VE) was assessed for the following case definition: clinical pertussis, laboratory-confirmed pertussis, epidemiologically confirmed pertussis (documented contact with a laboratory confirmed case). Immunization history was obtained by reviewing the child health record book. Effectiveness of a whole vaccination (four injections) and of a partial vaccination (one to three injections) were estimated as 1-(attack rate among vaccinated/attack rate among non-vaccinated). RESULTS: A whole immunization conferred good protection against pertussis with an estimated VE higher than 92% in three surveys, lower in the fourth survey (84%) in which antibiotic prophylaxis was set up very rapidly. A partial immunization conferred a mild protection (median: 60%). CONCLUSIONS: These results are consistent with a previous report about the effectiveness of this whole cell vaccine using the screening method in a hospital network survey in France. In the same way, a large efficacy trial in Senegal comparing it with an acellular bivalent vaccine estimated its efficacy at 96%. This high efficacy together with a satisfactory vaccine coverage leads to the current epidemiological profile of pertussis in childhood in France: majority of cases occurring before 6 months of age, limited outbreaks in school children, many of whom being unvaccinated or partially vaccinated.