Guidelines for red blood cell and plasma transfusion for adults and children
The risks of blood transfusion
The noninfectious risks
Physicians generally have better access to information on the noninfectious risks of blood transfusion than to that on the infectious risks. Therefore, noninfectious risks will be covered only briefly (Table 4).
Most acute hemolytic reactions due to ABO incompatibility can be avoided by appropriate testing before transfusion and use of procedures to prevent errors. Of 158 transfusion-related fatalities due to acute hemolysis reported to the US Federal Drug Administration in 197685, at least 86% were the result of process errors.141
In an analysis of fatal and nonfatal blood transfusion errors (transfusion of blood to other than the intended recipient or release of blood of an incorrect group) reported to the New York State Health Department from January 1990 to October 1991, 22% of errors occurred in phlebotomy and ordering, 32% within the blood bank and 46% during transfusion administration.121 The authors estimated a transfusion error rate of 1 in 12 000 transfusions and the death rate due to transfusion error was 1 in 600 000 units transfused. Further discussion of these potential adverse events is beyond the scope of this document; however, it is important to remember that several complications, in addition to acute hemolysis, may be fatal: delayed hemolysis, anaphylaxis, transfusion-related acute lung injury, graft-versus-host disease, posttransfusion purpura and iron overload. The interested reader is referred to recent excellent references.122,141
Furthermore, recent studies in animals and humans suggest that blood transfusions may cause immunosuppression.142,143
The infectious risks
The infectious risks described in this section apply only to nonfractionated blood components. Human blood can never be guaranteed to be completely free from infectious agents. However, risks change constantly over time, and the prevalence of the known infectious agents in collected blood has been decreasing as a result of more stringent donor-selection requirements and improved laboratory testing procedures.
The most reliable sources of data on the risk of infectious diseases related to blood transfusion are large, multicentre, prospective studies measuring seroconversion rates among repeat blood donors and the number of acquired infections in a large cohort of red-blood-cell recipients. There is relatively little information about infectious risks from blood collected in Canada; therefore, we have relied largely on data from the United States. The risks in Canada may be smaller because of a lower incidence of some blood-transmissible viruses in our donor population.144
When obtaining informed consent for blood transfusion, physicians may find that a comparison of the residual risks of infection from transfusion with the risks of death from various causes may be useful (Table 5). Similarly, estimates of risks related to the planned interventions (i.e., major cardiovascular surgery, general anesthesia) may help to put the risk of transfusion in perspective. Commenting on a report of transfusion-transmitted viral infections,145 Holland146 states, "Today, the serious risks from the blood components given to patients in transfusion are smaller than those from the underlying disease or the primary therapy."
Although responsibility for providing information to health professionals about the current risks of infection associated with transfusion of blood components probably lies with the blood collection and distribution agency, a hospital transfusion committee or equivalent body should assist in disseminating this information to health professionals and patients.
Viral contamination
Viral contamination of blood can occur only if the donor has viremia when the blood was being collected. In Canada, all units of collected blood are tested for serologic evidence of infection with human immunodeficiency viruses types 1 and 2 (HIV-1 and HIV-2), human T-cell lymphotropic virus type I (HTLV-I), hepatitis B virus (HBV) and hepatitis C virus (HCV). The HTLV-I assay will also detect most infections caused by HTLV-II.150 The residual risk of transfusion-associated infection (the likelihood of a screening test failing to identify the blood as infected) with one of the above viruses can be determined based on estimates of the number of donors giving blood during the infectious "window." The infectious window extends from a few days after the virus infects a person, during which there is rapid viral multiplication, until the earliest time that the screening test can detect its presence; if blood is collected during this period, it is presumed to be infectious to others through transfusion. The Retrovirus Epidemiology Donor Study (REDS),145 in the United States, of approximately 590 000 repeat donors who made 2.3 million donations from January 1991 to December 1993 has provided the following estimates of the chance of a unit of virus-contaminated blood remaining undetected: HIV, 1 in 676 000; HTLV (I and II), 1 in 641 000; HCV, 1 in 103 000; and HBV, 1 in 63 000.
The original HIV estimate of 1 in 493 000 in the REDS145 was based on current generation antibody tests, but did not include HIV-1 p24 antigen testing, which was introduced into routine blood donor screening in Canada and the United States in March 1996. The p24 antigen test was calculated to reduce the HIV window by 27.3% to about 19 days, which lessens the risk of an undetected but contaminated unit entering the supply to about 1 in 676 000 units. The REDS estimate is close to the upper limit of a range of 1 in 450 000 to 660 000 units reported in a study from the US Centers for Disease Control and Prevention and the American Red Cross.151 From a study conducted between April 1989 and March 1993, Remis and co-workers144 estimated an HIV risk of 1 in 913 000 units (95% CI, 1 in 507 000 to 1 in 2 050 000) for blood collected in Canada.
The REDS145 risk estimate for HBV infection of 1 in 63 000 is appreciably higher than the 1 in 200 000 cited by Dodd152; the lower rate was based on earlier information provided by the US Public Health Service.153
As many as 35% to 50% of Canadian blood donors test seropositive for CMV.154 Risk of serious outcome from transfusion-transmitted CMV infection is largely restricted to infants with very low birth weight (less than 1200 g); immunocompromised hosts, especially patients receiving bone-marrow transplants; and seronegative women in early stages of pregnancy where there is a risk of infection of the fetus. Recipients of solid-organ transplants, who received blood products not screened to eliminate units collected from CMV-positive donors, appeared to have little increased risk of CMV infection.154 The Canadian Red Cross blood centres have identified a group of CMV-seronegative blood donors who can supply blood components for at-risk patients on request. CMV is transmitted only by infected leukocytes; therefore, techniques that reduce the number of leukocytes in the transfused product, including use of a leukoreduction blood filter, will reduce the risk of CMV infection.155 Units of very fresh blood components carry a higher risk of CMV infection than products stored for several days.156
Hepatitis G virus (HGV) is considered to be the same virus as hepatitis GB virus, type C (HGBV-C), previously derived from another source. HGV RNA has been detected by polymerase chain reaction technology in 1% to 2% of volunteer blood donors in the United States. The virus is transmitted parenterally and has been found in people who have received a transfusion. Many people with serologic evidence of HGV are also infected with HCV and would be excluded as donors by laboratory assays for the latter. Although HGV infection may result in a chronic carrier state, there is no conclusive evidence that HGV causes acute or chronic hepatitis.130 Reports of a causal association between HGV infection and aplastic anemia in 2 patients have been challenged by Kao and colleagues,157 and the possible etiologic role of some strains of the virus in a few cases of fulminant hepatitis requires further clarification.158 Large-scale epidemiologic studies to define the epidemiology and natural history of HGV infection must await development of new laboratory technology.130
Infections caused by contamination of blood components with other organisms such as hepatitis A virus and parvovirus B19, for which blood donors are not serologically screened, have been documented but estimates of rates of transfusion-associated infections are not available. The risk of serious clinical consequences of parvovirus B19 infection is greatest for patients with anemia that results in increased red blood cell production (e.g., sickle-cell disease) and possible development of transient aplastic crisis; for people with immunodeficiencies who may develop severe, chronic anemia; and for nonimmune pregnant women, in whom the outcome may be intrauterine death associated with hydrops fetalis. The risk of fetal death is less than 10% after proven maternal infection in the first half of pregnancy and may be negligible in the second half. Congenital anomalies have not been reported among newborn infants in association with parvovirus B19 infection.159
Bacterial contamination
Bacterial contamination of blood donor products may result from exposure to skin bacteria introduced through venipuncture when blood collection is started; use of blood collection and administration apparatus contaminated at source by the manufacturer; manipulation of the blood product by blood bank or hospital staff administering the transfusion; or by bacteremia, usually asymptomatic, in the donor at the time of blood collection (e.g., Yersinia enterocolitica).
The number of contaminating bacteria increases with the duration of storage of red blood cells or plasma before transfusion, thus increasing the possibility of clinical signs of infection in the recipient. Storage at room temperature enhances the growth of most bacterial pathogens, but Yersinia multiplies readily at storage temperatures as low as 4°C.
Clinically significant bacterial contamination of red blood cells is less than 1 in a million units transfused.160 The risk of death from bacterial sepsis has been estimated at 1 in 9 million units of red blood cells transfused.161 Bacterial contamination of blood components accounted for 16% of transfusion-associated fatalities reported in the United States between 1986 and 1991; 28% of these were associated with transfusion of red blood cells. The risk of bacterial contamination is not reduced by the use of autologous blood transfusion.162
Transmission of syphilis by transfusion in Canada has been virtually eliminated by careful donor selection and serologic testing of each unit for markers of syphilis infection.
Parasitic contamination
Parasitic contamination of blood can occur only if the donor has a parasitemia, usually asymptomatic, at the time of blood collection. Donor selection criteria based on a history of recent travel to, or former residence in, endemic areas greatly reduce the possibility of collecting blood from those capable of transmitting malaria, Chagas' disease or leishmaniasis. Reported cases of transfusion-associated Chagas' disease are extremely rare in the United States and Canada. Transfusion-associated cases of malaria have occurred in recent years in Canada, and the residual risk of receiving a unit of red blood cell concentrate contaminated with malaria parasites is estimated at 1 in 400 000 units (Anthony Giulivi, MD, FRCPC, Canadian Red Cross Society, Ottawa: personal communication, 1996).
Creutzfeldt-Jakob disease
This rapidly progressive, fatal, degenerative neurologic disease is believed to be caused by an infectious protein agent called a prion. People considered to be at risk of Creutzfeldt-Jakob disease (CJD) as a result of past corneal or dura mater grafts, injections with growth hormone or gonadotropins derived from human brain extracts or having a history of CJD in a first-degree blood relative are permanently excluded from blood donation. Blood components and plasma derivatives from a person subsequently found to be at risk of CJD or subsequently diagnosed to have CJD are immediately withdrawn from further distribution. These precautions have been undertaken even though a case of transfusion-related CJD has never been confirmed in humans.163
A definite history of blood transfusion was found in 16 of 202 patients with CJD; this rate was similar to that in a matched control group.164 An analysis of multiple-cause-of-death mortality data for 197994 in the United States revealed 3642 death certificates on which CJD was listed as a cause of death; none listed thalassemia, sickle-cell disease or hemophilia -- diseases associated with high exposure to blood products. A 1995 US study uncovered no hemophilia patients with CJD.165 The growing epidemiologic data in the United States strengthen the conclusion that the risk, if any, for transmission of CJD by blood products is extremely small and, at present, appropriately regarded as theoretical.165 The Laboratory Centre for Disease Control, Health Canada, has launched an active surveillance program of CJD and is undertaking extensive epidemiologic studies to monitor any possible relation between transfusion and CJD.
Recommendations regarding the infectious risks of blood and plasma
- Current, accurate information pertaining to the infectious risks of red blood cells and plasma should be accessible to physicians, other health care providers and consumers.
Level of evidence: N/A
- Local committees responsible for blood transfusions for an institution or a regional health authority should ensure that accurate information about the infectious risks of blood transfusion are disseminated to health care providers.
Level of evidence: N/A
