Non-infectious Serious Hazards of Transfusion

Safe transfusion therapy is a basic requirement for advanced medical care. In order to establish priorities for improving safe transfusion for patients, it is essential to distinguish transfusion safety from blood safety. Blood safety refers to the safety of the product. In contrast, transfusion safety refers to the safety of the overall process of transfusion from donor to recipient. Blood safety is not matched by increases in transfusion safety. Enormous progress has been made in blood safety during the last few decades. Donor restrictions and high performance viral screening assays have virtually eliminated viral transmission by blood transfusion in the developed world. For example, the risk of hepatitis or HIV from transfusion has declined by approximately 10,000 fold. In contrast, during the same period there has been little change in the risk of non-infectious hazards. Some result from medical errors made during the collection of a patient sample or during blood administration. Data suggest that the rate of medical errors occurring in hospitals is increasing and a labor shortage in hospital laboratories also decreases patient safety. For example, a recent study from a prestigious university program documented that from 1993 to 1999 the number of errors rose from 1238 to 2052 per year. It is particularly alarming that errors involving patient samples quadrupled from 112 (10% of total errors) in 1993 to 434 (20% of total) in 1999. The Institute of Medicine report, To Err Is Human, has called attention to the enormous morbidity and mortality associated with hospital-based errors. These errors have been largely ignored by government agencies focused on blood safety. For example, the Canadian Krever Commission selectively applied the precautionary principle only to infectious risks of transfusion, but not to the very real noninfectious hazards of transfusion. Thus, while blood product safety has been a remarkable achievement, emphasis on product safety diverted attention from transfusion safety. Hemovigilance programs and the risk of transfusion. Hemovigilance programs are national systems for reporting adverse events and provide one objective means to assess current risks of transfusion. In reports of adverse occurrences in the United Kingdom, mis-transfusion accounted for over 50% of adverse events and non-infectious hazards of transfusion accounted for over 95%. Similar data generated from hemovigilance programs in France and Canada suggest that innovations to address non-infectious hazards should be given high priority. Viewed from the perspective of risk per individual unit, noninfectious hazards overwhelm current infectious risks. Figure 1 shows current estimates of the risks of an individual unit of blood using the Paling scale. These data demonstrate two important findings. First, the confidence interval of the risk estimate is much more precise for infectious hazards compared with noninfectious hazards. The precision of infectious risk estimates has resulted from studies conducted in the last decade. Similar studies for non-infectious hazards have not been conducted. Secondly—and more importantly—the graph shows the extent to which patients are at greater risk from non-infectious serious hazards of transfusion. For example, estimates of the perunit risk of mis-transfusion errors may exceed the risk of viral infection by as much as 10,000 fold. Examples of non-infectious hazards of transfusion: Mis-transfusion of blood. Mis-transfusion can be summarized as a failure to give “the right blood product to the right patient at the right time for the right reason.” At its worst, mis-transfusion results in major ABO incompatible transfusions. Despite dramatic improvements in overall medical care in the past half-century, the morbidity (renal damage) and mortality associated with ABO hemolytic transfusion reactions has not improved much over that observed decades ago. For example, transfusion errors comprised 2.7 % of all sentinel events reviewed by the Joint Commission on the Accreditation of Hospitals from January 1995 through March 2002, although these and other data from the US based on passive reporting may underestimate the true frequency of mistransfusion. An active audit of transfusions at three university hospitals in Belgium observed numerous unreported mistakes. Overall, the incidence of serious error was 1 in 400 units and the rate of reported errors underestimated the true rate by 30 fold. Transfusion-related acute lung injury (TRALI). TRALI is an immune-mediated lung injury syndrome, which in its worst form is life-threatening and indistinguishable from adult respiratory distress syndrome. The actual incidence is uncertain and many cases undoubtedly are attributed to other causes. Over a two-year period in a general hospital, Clarke et al reported that 46 of 2,430 transfusions of platelets (2%) were associated with respiratory reactions. A more recent study observed frequent oxygen desatu-