|Home | About | Journals | Submit | Contact Us | Français|
In the past few years there has been increasing concern about blood transfusion safety. Avoidable transfusion errors, mostly in patient identification, remain a serious cause of injury and death. There is also heightened awareness of the risk of transmission of viral and bacterial infections. Of particular concern in Britain is the (theoretical) possibility of transmission of variant Creutzfeldt-Jakob disease.
This review puts these risks in perspective (table) and describes the new measures that have been introduced to improve blood safety. It also describes changes in attitude and practice that will affect users of blood in all disciplines, including general practitioners advising patients of the pros and cons of transfusion. Finally it emphasises the need for careful education and training of all those involved in blood prescribing and blood component administration.
Our review is based on information from the annual reports of Serious Hazards of Transfusion (www.shot.demon.co.uk/), the guidelines of the British Committee for Standards in Haematology (www.bcshguidelines.com/), and the chief medical officer's second “Better Blood Transfusion” meeting (www.doh.gov.uk/bbt2). We also cite relevant recent publications by leading clinicians and scientists.
Avoidable transfusion errors remain an important if uncommon cause of death and injury. In the United States fatal misidentification errors are estimated to occur in 1 in 600000 to 1 in 800000 transfusions and non-fatal errors occur in 1 in 12000 to 1 in 19000 cases.2,3 UK data from the Serious Hazards of Transfusion (SHOT) reports suggest an error incidence of 335 per 5.5 million units of red cells transfused. The most commonly reported adverse event, “incorrect blood component transfused,” accounted for nearly 70% of reports in 1999-2000.4 Incompatibility in ABO blood groups was reported 97 times and led directly to four deaths and 29 cases of immediate major morbidity.
After the second SHOT report, updated UK national guidelines to minimise the risk of giving the wrong blood were published.5 In the past two years many hospitals have introduced hospital-wide “adverse incident reporting” schemes to identify and analyse such incidents and “near misses.” Transfusion errors feature prominently among these incidents (personal communication, F Regan). Existing adverse clinical incident reporting schemes will probably soon feed into a central UK reporting scheme managed by the National Patient Safety Agency to generate national information and recommendations. Recognition that educating staff and implementing robust hospital transfusion protocols are needed to prevent errors has resulted in these factors being incorporated in the Clinical Negligence Scheme for Trusts. However, training all staff involved in blood administration or taking samples for cross matching, including locum and agency staff, will be difficult without adequate resources.
Internationally, new information technology systems are being developed to design error out of the transfusion process.6 These are based on a unique barcode on each patient's wristband, which is transferred on to the patient's cross match blood samples and transferred to each unit of blood prepared for that patient. This barcode is matched electronically with the patient's wristband before administering blood (fig (fig1).1). Pilot studies are currently assessing the feasibility of these systems in various settings including day wards, presurgical admission clinics, and inpatient wards.
Safety measures to minimise the risk of transmitting known infections through transfusion include donor selection and exclusion, testing of donor blood, and post-collection processing such as leucodepletion and viral inactivation (see below). National haemovigilance schemes to monitor adverse transfusion events have been introduced in many countries,4,7,8 and EU-wide data are being collated by the European Haemovigilance Network. Similar systems exist in the United States and Canada.
Despite these measures, the possibility of transmission of new infectious agents, including variant Creutzfeldt-Jakob disease (vCJD), remains. Although there is no evidence of vCJD transmission in humans, concern has been provoked by a study in which one of 19 asymptomatic sheep, 318 days after being given 5 g of cow brain infected with bovine spongiform encephalopathy (BSE) in their feed, seemed to transmit BSE to a second sheep via a 400 ml blood transfusion.10,11 Although no other studies have been published to validate this finding, steps have already been taken in Britain to reduce the possible risk of vCJD transmission by transfusion (box (boxB1B1).12 In addition, the Department of Health's Advisory Committee on the Microbiological Safety of Blood and Tissues for Transplantation is considering excluding blood donors who themselves received transfusions between 1980 and 1996. The problem with this is it would result in a loss of about 10% of donors, and, without a corresponding reduction in blood use, blood stocks would be severely jeopardised. Furthermore, the blood supply would probably be further reduced if a blood test for vCJD becomes available.13
Several companies are working to produce a screening test for vCJD, and one is likely to be available within two years. Once it is, the National Blood Authority will be under pressure to introduce it. (In recent litigation in relation to the transfusion of hepatitis C the National Blood Authority was found at fault for supplying a defective product, and the avoidable delay in implementing an available hepatitis C test was highlighted.14) Anonymous testing will not be an option: under EU law, donors must give consent for all tests performed on their blood and must be informed of any test results on the which the national blood authority acts (for example, discards their blood). It is likely that many donors will not agree to be tested, as the burden of knowledge will affect not only their health and happiness, but could affect availability of life insurance policies. Importing blood from BSE-free countries may seem attractive, but, as most countries face periodic blood shortages, it is unlikely that sufficient blood would be available to replace the UK blood supply of around 2.7 million units of red cells a year.
Considerable variation in transfusion practice for elective surgery is well documented (fig (fig22).15 Reducing unnecessary exposure to blood components by blood saving measures is particularly important in healthy patients undergoing elective surgery (box (boxB2).B2). A recent publication for anaesthetists summarises good transfusion practices in surgical patients.16 Implementation has been problematic, however, as until recently blood has been perceived as a safe and unlimited resource, and it has been difficult to secure funding for blood saving measures.
About half of all blood transfused in the United Kingdom is to surgical patients (National Blood Service internal audit). To reduce the amount of blood used in elective surgery, detailed planning at each stage of patient care is required (box (boxB3,B3, fig fig3).3). Although the cost of the blood component may be saved, other costs may be incurred and there may be no overall saving in the short term. Long term savings relating to the potential cost of transfusion transmitted infection, immunomodulation (long term mild immune suppression which occurs in recipients of blood components and can result in poorer outcome17), and litigation may be substantial but are difficult to quantify.
Implementing strategies to reduce the requirements for blood transfusion requires effective teamwork, adequate resources, and a clear understanding of the rationale for it. Blood substitutes, such as haemoglobin solutions and perfluorocarbons, are in phase III clinical trials, but their short half lives may limit usefulness.18,19 Another approach to reducing unnecessary transfusion would be to enforce, either locally or nationally, a policy of blood components being prescribed only by senior doctors.
Over the past three years, specialist practitioners of transfusion have been appointed in over 40 UK hospitals, echoing similar developments in Europe and the United States.8,9 Most are senior nurses, but some are doctors or biomedical scientists. These posts have been created to implement recommended policies to reduce inappropriate prescribing of blood components.20 Although the cost of employing specialist transfusion practitioners has deterred some trusts, it has been found repeatedly that the savings from reducing inappropriate prescribing of blood products exceed the cost of employment.
The main role of the specialist transfusion practitioners is to educate staff and patients about the pros and cons of blood transfusion and to support the development and evaluation of transfusion protocols and guidelines. They also facilitate audit and implement strategies to improve blood ordering and administration.21 Where appropriate, practitioners may be directly involved in near patient testing and cell salvage techniques.
Blood components are becoming safer as more sensitive screening tests for viruses are introduced. In the United Kingdom all cellular blood components have been leucodepleted at source since November 1999 to reduce the potential transmission of vCJD, thought to be facilitated by B lymphocytes.22 Leucodepletion also reduces transmission rates of other cell associated viruses such as cytomegalovirus.23 The recent introduction of a nucleic acid test for hepatitis C in fresh frozen plasma, blood, and platelets24 has reduced the “window period” from 70 days (for antibody testing) to 13 days, and the chance of transmission by a unit of blood from 1 in 250000 to 1 in 3 million.22
To reduce risks further, viral inactivation steps, routinely applied to pooled fractionated products such as albumin or immunoglobulin solutions, could now be applied to fresh frozen plasma and possibly cellular components.22
Pooling of plasma from over 1000 donors is required for solvent detergent treatment of fresh frozen plasma and fractionated products, for efficiency of processing and product standardisation. Pooling theoretically allows contamination of the entire pool by an infectious agent from one donor. Although the treatment kills enveloped viruses such as hepatitis B and C and HIV, not all non-enveloped viruses are affected (such as hepatitis A and parvovirus). Serological and polymerase chain reaction testing of the plasma pools is also carried out, but not all known agents are tested for, and some transmissions of parvovirus have occurred.25 Use of solvent detergent treated plasma is widespread, and in some European countries the use of untreated plasma is banned. An alternative is methylene blue treatment, which can be applied to single units of plasma. This inactivates a broader spectrum of viruses but is more costly and time consuming. Methylene blue is also potentially more toxic.26
The UK Advisory Committee on Microbiological Safety of Blood and Tissues is currently considering for which groups of patients fresh frozen plasma from UK donors should be virally inactivated. It is also looking at possible alternative sources of fresh frozen plasma. Methylene blue treatment of fresh frozen plasma, from UK donors is being introduced from May 2002 for children and infants born after 1 January 1996, the date when vCJD was officially excluded from the human food chain in Britain.
Solvent detergent and methylene blue treatments have no effect on bacteria or prions—there is no known suitable way of inactivating prions, which are resistant even to extremes of temperature. However, bacterial contamination of blood components, especially of platelets, is a more important cause of mortality and morbidity from blood transfusion than is viral transmission.4 A third method of pathogen inactivation is therefore being considered that not only inactivates all viruses but also kills bacteria, parasites, and lymphocytes. Psoralen S-59 and ultraviolet light are used together to treat individual platelet concentrates in the Helinx system, which cross links DNA and RNA.27 Another psoralen, S-303, is in development for use in red cell concentrates. Although expensive and labour intensive, this system could inactivate all potential pathogens except prions. In addition, this treatment would make it unnecessary to irradiate blood components to prevent transfusion associated graft versus host disease as the donor lymphocytes responsible would be killed. In the future it may become the pathogen inactivation system of choice, unless it is overtaken by new developments.
Both authors also work at the National Blood Service, North London Centre, Colindale, London. We thank Dr Mahes de Silva and Dr Kevin Barraclough for their helpful comments on reviewing the manuscript and Ms Carmel McGinn for preparing the manuscript.
Competing interests: CT has a transfusion data manager funded by Ortho-Biotech, which manufactures erythropoietin, and has received fees and travel costs from the company for speaking at a symposium.