Quality of transfusion practice beyond the blood transfusion laboratory is essential to prevent ABO‐incompatible death

The articles on improving safety in blood transfusion in this issue highlight two important areas. The paper by Lau et al. (pp. 121±124) describes an ingenious wristband with a barcode system for a cost of around 33p per patient. They also describe a scanner to replace the need for two people to perform the double check of identity of the patient and the donations or products to be transfused. Their study is rather small and deserves further evaluation; certainly their cheap scanner (£250) would prove safer than recent contentious use of just one person to perform the double check prior to transfusion. However, this system, like all the others, does not guarantee getting the wristband on the correct patient if two patients are bled prior to labelling. In addition, wristbands may be removed in the operating theatre and some day-patients refuse to wear them. The article by Cummins et al. (pp. 117±120) on the use of addressograph labels highlights current problems in which some hospital transfusion committees agreed not to follow the BCSH guidelines on the administration of blood and blood components and the management of transfused patients (BCSH, 1999) because they found too many minor errors when sample labels were handwritten. The use of addressograph labels on samples used for compatibility tests may result in fewer errors of a minor nature, but some serious errors of the incorrect ABO group had occurred that could have led to ABO-incompatible transfusion. Other attempts have been made to cover the lack of a quality system, including bedside testing (see review by Baglin et al., 1990), computerized bedside transfusion system (Jensen & Crosson, 1996) and a mechanical barrier system (Wenz & Burns, 1991). However, these are either too expensive or too complex. Blood Donor Centres in the UK are monitored to high quality standards by the Medicine Control Agency and the ABO and RhD type of donations in the UK has been guaranteed since July 1998. Hospital Blood Transfusion laboratories are subject to CP Accreditation and Health Quality Service Accreditation but these cover only those areas controlled by Pathology and are management rather than a quality system to assure technical competence (Phillips et al., 1994). Recent surveys (McClelland & Phillips, 1994; Williamson et al., 1998) con®rm that ABO-incompatible death and morbidity still occurs in NHS Hospitals and con®rms these are mainly due to errors beyond the laboratory. Correct patient identi®cation before sample collection and the administration of blood are key areas of safe transfusion practice. It is also likely that the number of reports of ABO incompatible deaths or morbidity reported to SHOT are underestimated. This may be due to nonreporting or misdiagnosis in the unconscious patient. It is important to realize that sera from group A and group B recipients contain much less anti-B or anti-A, respectively, compared with the anti-A/anti-B levels found in the sera of group O recipients. Consequently, ABO-incompatible transfusion of group A red cells into group B recipients or group B red cells into group A recipients can pass without observable effects, leading to a dangerous illusion of safety. Transfusion of group A or group B donations into group O recipients is, however, far more dangerous as the high levels of incompatible anti-A or anti-B may cause massive reactions leading to disseminated intravascular coagulation (DIC). DIC in this situation may present as a coagulation problem rather than an ABO-incompatible transfusion leading to serious delays which may have lethal consequences. The cause of most of failures in transfusion practice is the absence of a full quality system. A quality system that brings the identi®cation of the patient and the patient's sample and the administration of blood and subsequent monitoring of the transfusion into the quality loop is essential for safe transfusion practice. This quality loop involves different staff groups and has numerous managerial interfaces that have to be effectively managed. It is these managerial interfaces that present the main obstacle to eliminating these errors. The number of interfaces in the transfusion loop requires implementation of a quality system that recognizes them and deals with them. The problems include: X lack of suf®cient awareness of the importance of patient identi®cation and sample identi®cation, and the considerable variation in identi®cation requirements in the different Trust Hospitals; X management of phlebotomy services. Some 25% of hospitals have moved the responsibility for phlebotomy Transfusion Medicine, 2000, 10, 95±96