Article Text

Download PDFPDF

Blood transfusion
  1. P H B Bolton-Maggs1,
  2. M F Murphy2
  1. 1Manchester Royal Infirmary, UK
  2. 2National Blood Service and Oxford Radcliffe Hospitals, and Senior Clinical Lecturer in Blood Transfusion, University of Oxford, UK
  1. Correspondence to:
    Dr P Bolton-Maggs
    Department of Clinical Haematology, Manchester Royal Infirmary, Oxford Road, Manchester M13 9WL, UK;

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Risks and benefits

Transfusion of blood and blood products to infants and children is not without risk. In recent years, the risks of blood transfusion have been highlighted by the reports of the Serious Hazards of Transfusion scheme (SHOT), which collects data on adverse events related to transfusion. There continues to be concerns about transfusion transmitted infection, most recently in relation to variant Creutzfeldt-Jakob disease (vCJD). At the same time, new guidelines have recently been published to optimise transfusion practice for children.1,2 This editorial addresses some of the risks and benefits of transfusion therapy in children, concentrating on aspects that may be less familiar to paediatricians.

Since the development of the blood transfusion services during and after the Second World War, we have perhaps come to take the provision of this valuable resource for granted. The advent of plastic bags enabled separation of a unit of blood into component parts—the ready availability of platelet concentrates allowed intensification of chemotherapy regimens and contributed towards improved survival and reduced morbidity. The National Blood Service currently collects about 2.5 million donations of blood from voluntary unremunerated donors in the UK every year, but only about 3% of the blood and blood products are transfused to children under 16 years of age.3 Most blood is transfused to people more than 60 years of age whose life expectancy is frequently reduced due to their underlying diseases. The consequences of transfusion transmitted infections, many of which have a long latent period before clinical expression, to neonates and young children with a normal life expectancy will be much more significant than to the elderly. In UK hospitals we expect that blood will be available for any child in need, and that its transfusion is safe and effective. Quite apart from a consideration of the risks of transfusion to children, what do we know of the benefits? A closer examination of our practice indicates that little of our paediatric transfusion practice is well grounded in satisfactory evidence by today’s standards.


Transfusion transmitted infection is probably the best publicised of the hazards of transfusion, certainly in the minds of parents who may ask about HIV transmission. Recent figures made available from the National Blood Service show that the risk of HIV transmission in the UK has been reduced to an almost negligible level (to less than 1 in 8 million, less than the risk of being struck by lightning). Since the introduction of hepatitis C antibody testing of donors in 1991, and more recently the testing of donor minipools (48 donations per pool) by methods designed to detect viral DNA, the risk of contracting hepatitis C virus has decreased to less than 1 in 30 million. The risk of hepatitis B transmission is 1 in 900 000. So in terms of viral transmission, blood transfusion is much safer than in the 1960s when the risk of contracting hepatitis was as high as 1 in 5 associated with the use of paid donors in the USA. Blood Services continue a vigilant search for methods to prevent transfusion transmitted infection as new infectious agents appear in different parts of the world; for example, the National Blood Service is currently revising its donor questionnaire in relation to recent travel to take account of the possible risks of West Nile Virus and severe acute respiratory syndrome (SARS). West Nile Virus has recently emerged as a significant problem in the USA with 23 cases of infection confirmed transmitted by transfusion. As a result, the blood collection industry in the USA initiated a voluntary withdrawal of stored units of FFP and cryoprecipitate that might contain the infection—as there was no test available at that time. Within a few months this outbreak had been fully documented and a new nucleic acid based assay developed for donor screening.4 A high level of surveillance continues in areas at risk for extension of the epidemic as the virus spreads to new areas.

In the UK very serious consideration is being given to the possibility of transmission of the agent responsible for vCJD. The risk of transmission of this agent (which is an abnormal prion protein) is currently unknown, but transmission of bovine spongiform encephalopathy (BSE) by blood transfusion has been shown in sheep.5 It is possible that the agent is present in UK blood donors, and that recipients might be infected. To this end, the Department of Health and the National Blood Service are considering how the risk can be assessed, and reduced. The introduction of leucocyte depletion of blood components, implemented in the autumn of 1999, may reduce the risk by removing lymphocytes; in animal models where infectivity has been found in the peripheral blood, a large proportion has been associated with white blood cells.6 Also since 1999, plasma for fractionation into blood products such as intravenous immunoglobulin and factor VIII concentrates has been sourced from the USA, a country with no documented cases of BSE. For children born after 1 January 1996, fresh frozen plasma (FFP) will be sourced from plasma (from unremunerated donors) imported from the USA within the next few months on the basis that exposure to BSE from food was eliminated by 1 January 1996. UK sourced FFP for this group is already being treated with methylene blue to inactivate viruses. Bacterial infections in blood products are very rare, but can be lethal—over the first six years of reporting by SHOT, bacterial contamination has accounted for 26/40 transfusion transmitted infections reported, including six deaths. Platelet transfusions were the commonest source of bacterial infections, responsible for 22/26 of the cases of bacterial contamination reported to SHOT; this is probably related to their storage temperature of 22°C. The platelets were three or more days old in 21/22 cases; the shelf life is only five days. Bacterial contamination may be an under-recognised hazard as fever in sick neutropenic patients may be attributed to other causes. Consideration is being given to methods for reducing the risk of bacterial infection, including testing of platelet concentrates and the introduction of methods for “pathogen inactivation” of blood components, which would also prevent transfusion transmitted bacterial infection.

There are several other potential hazards of transfusion which cannot be predicted in advance, such as acute and delayed transfusion reactions, post-transfusion purpura, transfusion associated graft versus host disease (TA-GVHD), and transfusion related acute lung injury (TRALI) (table 1).7 Many of these may be associated with relatively minor morbidity and mortality, but TA-GVHD is uniformly fatal, and TRALI in children has had a particularly high mortality. It is important to take steps to provide relevant groups of patients with gamma irradiated blood to avoid TA-GVHD, and to have a clinical awareness of TRALI so that appropriate treatment may be initiated. The reader is referred to the new guidelines, which discuss transfusion requirements for children in specific clinical situations in detail, and give advice about how to avoid some of the above hazards.1 The Handbook of transfusion medicine8 gives further information about these and many other aspects of transfusion and can be found on the web at

Table 1

Overall mortality/morbidity figures by fully analysed questionnaires 1996/97–2001/02 (n = 1630) (from the SHOT report 2001–027)

However, it is a sad reflection on our daily practice that the greatest risk associated with transfusion remains that a patient will receive the wrong blood (a risk of about 1 in 25 000 units transfused).9–11 The SHOT scheme, established in 1995, has shown this to be the largest group of adverse events. The scheme is a confidential enquiry for the reporting of serious complications of blood transfusion and near miss events in the UK. The cumulative results of six years reporting show that 64% (1093 of 1711 incidents) were in this group, and there is no evidence of any reduction year on year (table 2).7 Analysis of data from children up to the age of 18 shows that a higher proportion occur due to incorrect components transfused (80%), but the actual numbers of incidents in children are small. Fortunately the harm done is usually minimal (table 1) but death or major morbidity does occur due to ABO incompatible transfusions (a total of seven deaths was reported in six years to SHOT). This should alert us to improve our practice, and invest in resources to improve the education and training of the many staff involved in the transfusion process. New technology is becoming available to improve the performance of the clinical transfusion process, including sample labelling and the bedside clerical check.12,13 Several technology solutions are in various stages of development, and include wireless hand held portable digital assistants, advanced bar coding, radiofrequency identification, and imbedded chip technology. Such devices for transfusion safety are promising, but need to undergo clinical trials to show their effectiveness and ease of use in routine practice. They will also need to be integrated with other hospital procedures requiring patient identification in order to be financially acceptable to hospitals.

Table 2

Adverse events reported during the six reporting years 1996/97 to 2001/02 (from the SHOT report 2000–017)

Publication of the results from SHOT contributed to meetings led by the chief medical officers and the production of two health circulars, Better blood transfusion, published in 1998 and 2002.14,15 In accordance with clinical governance, these set standards and action points for hospitals. The first circular (HSC 1998/22414) required all hospitals to have a transfusion committee to oversee all aspects of transfusion and participate in the SHOT scheme. Trusts were encouraged to agree and disseminate local protocols for blood transfusion based on guidelines and supported by in-house training. Trusts were also encouraged to review indications for transfusion and to ensure that the indication for transfusion was clearly documented in the patient’s clinical case notes. Alternatives to transfusion (for example, autologous transfusion, cell salvage procedures) should be explored. The second circular (HSC 2002/009,15 which can be viewed at made further recommendations about the appropriate use of blood and blood products. A number of actions were to be taken by December 2002 and others by April 2003 (for example, to ensure that transfusion protocols are in place, and monitored, to ensure that education and training takes place for all health care staff involved in transfusion). The implementation of these recommendations requires an increase in resources in Trusts, both in the allocated sessions of consultant haematologist time, and the appointment of specialist practitioners of transfusion who will assist in teaching, training, and audit. These individuals together with the hospital blood bank manager will make up the Hospital Transfusion Team. A formal education programme can improve the safety and effective use of transfusions.16

Transfusions should only be given when clearly indicated, and with appropriate monitoring. Doctors must be able to justify the exposure and risks. Parents may well enquire if they can give blood for their own child—“directed donation”. This request is usually triggered by an unrealistic fear in the parent of the risk of transmission of infections such as HIV. There are occasional indications (for example, a very rare blood group, special HLA or other rare antigen compatible platelets), but in general this is discouraged for several reasons—directed donors have been shown overall to have a higher rate of positivity for viral markers, they are likely to be emotionally involved, and there are definite immunological risks.17 No blood service endorses the practice and studies have shown that directed donor programmes are not justified.


There is no doubt that transfusion of blood and blood products is necessary and appropriate in a number of situations in paediatric practice, and the revised national paediatric transfusion guidelines will assist the clinician.1 The new guideline updates the first version published almost a decade ago,18 but a major problem is the lack of a satisfactory evidence base for many of the recommendations. There are few randomised controlled clinical trials relating to transfusion in paediatrics and there is a need for the design and funding for more. In paediatric practice it is clearly difficult to choose end points for such studies. However, it is important for paediatricians to critically examine their transfusion practice and ensure that the clinical indications for transfusion are clearly documented in the case notes. Wherever possible, protocols should be in place enabling regular audit. Subspecialties could develop agreed standards within their own area against which audit could be carried out; for example, there are clear differences of views between neonatologists concerning the ideal blood product for exchange transfusion, and the threshold for top-up transfusions.

The availability of blood and platelet transfusion support has permitted increasingly more intensive chemotherapy regimes to be used in malignant disease at all ages,19 and recent randomised controlled trials in adults have shown that the threshold for prophylactic platelet transfusions can be safely lowered from 20 to 10×109/l.20 Similar trials are needed in this and other areas of neonatal and paediatric transfusion practice.

Children with haemoglobinopathies such as beta thalassaemia major, or with red cell aplasia are dependent on regular transfusions. These children should have a more extended blood group undertaken prior to the first transfusion in order to minimise the development of red cell alloantibodies, and blood should be regularly selected to match for the most immunogenic antigens. Premature neonates are a particularly highly transfused group at least in part due to iatrogenic blood loss. It should not be forgotten that severe immune deficiency, and thereby the first evidence for the transmission of HIV by transfusion, was shown in this group.21 Encouraging changes in practice have been shown with re-evaluation of the transfusion thresholds and the use of multiple satellites from one donor in order to reduce donor exposure. However, there are still differences of opinion about best practice, resulting in different recommendations for neonates between guidelines from the USA22 and the UK.1 Other changes in practice are evolving. A recent systematic review has shown that fewer infants require exchange transfusion for haemolytic disease of the newborn when high dose intravenous immunoglobulin is used.23 Review of the variable practice in different paediatric cardiac surgery units across the UK has shown a need for further studies and comparisons, as in adult practice. To this end a new national group (Haematologists Associated with Cardiac Surgery, HACS) has been formed to consider some of these issues and to develop national studies and trials.

There are many other areas where paediatricians could examine more closely the indications for transfusion, including for surgical procedures. In adult practice wide variations are found in transfusion practice related to the same operation (hip and knee replacements) between different hospitals.24,25 This kind of comparative audit has been very successful in changing practice.26,27 Perhaps the British Association of Paediatric Surgeons and other subgroups of paediatricians whose patients are frequently transfused could initiate similar comparative studies across the UK.

The new guidelines are to be welcomed, not least as an encouragement to renewed efforts to carry out clinical trials, so that our practice can be more firmly rooted in adequate evidence. Now is the time for every paediatrician to critically re-evaluate his or her local transfusion practice. Decisions to transfuse children should be made by doctors with adequate experience and training. Wherever possible, the reasons for transfusion should be fully discussed with the child and parents, appropriate written information given to support the discussion, and the decisions, together with the reasons for them, carefully documented in the case notes. The effectiveness and adverse effects of any transfusion should also be recorded. These data constitute the minimal essential data set for every transfusion episode, and should be standard practice in every hospital.

A further important reason for critical evaluation of our transfusion practice is that in future supplies may be restricted. For example, future strategies for prevention of transmission of vCJD by blood transfusion may include restricting blood donation to those individuals who have never received a transfusion in the past, and the introduction of a test for the prion protein associated with vCJD with exclusion of donors found to be positive. The first of these strategies would probably reduce the donor population by 5–10% and would be likely to lead to an immediate shortfall in the blood supply within a few days. Doctors nationwide are being asked to review their transfusion policies and to see where transfusions can be reduced or avoided. A blood test is not yet available, but may become so within the next 3–5 years. It is unknown what impact this might have on the blood supply. The National Blood Service scenario planning is working with a worst case of a reduction to 50% of current donors. Many donors may not wish to donate knowing they are to be tested for an agent when the implications of a positive test are not known. Experience with hepatitis C testing, infections, and the law, has shown that if the existence of a defect in blood is known or should have been known, then the NBS continues to produce blood and blood products at its own risk.28 In view of the uncertainty of the safety of blood transfusion clinicians should carefully weigh the necessity for each transfusion, and intensify the search for alternatives to transfusion.29

Many other transfusion guidelines are available and can be accessed from links at,, and

Risks and benefits



  • Both authors are members of the Appropriate Use subgroup of the National Blood Service Blood & Tissues Safety Assurance Group