The risk of transmitting viral infections by transfusion today is quite remote. The many, sensitive, diagnostic tests in place, when applied to the blood of volunteer, unpaid (unremunerated), unpressured donors who are also carefully evaluated at the time of donation, make blood and blood component transfusions very safe. A number of sensitive laboratory tests are performed on each unit of donated blood and plasma to reduce the risk of transmission of hepatitis viruses and retroviruses from asymptomatic donors to transfusion recipients. With the tests, we hope to catch otherwise undetectable individuals who may be carrying these viruses yet appear healthy and deny risk factors for their carriage. However, the laboratory tests in use in blood banks were designed to aid in the diagnosis of patients with viral diseases. Therefore, a reactive test, even if reproducible, on a sample from a healthy blood donor is more apt to be falsely than truly positive. An ideal microbiologic test is one which is one hundred percent sensitive, i.e., it will identify every person with an infectious disease (including asymptomatic carriers). In addition, a perfect test would have one hundred percent specificity, i.e., it would not be reactive in anyone without the infectious agent. The decision point or ''cutoff'' for an ideal test would be above the (negative) results for all normal and uninfected samples, but below that for all (positive) infectious ones. In reality, there is an overlap between some of the results on normals and those on diseased individuals, including persons who are carrying an infectious agent. When we try to obtain maximal sensitivity,e.g., to detect all asymptomatic carriers of a virus, the assay cutoff is set very low for tests applied to blood donors; but this approach will compromise the specificity of a test. The net effect is that many normal people donating blood are said to have ''abnormal'' test results which, among other things, necessitates the loss of their blood and plasma. In addition, we must follow up the reactive results by enzyme linked immunoassays (EIA or ELISA) or radioimmunoassays (RIA) used to screen or preliminarily test blood from donors with supplemental or confirmatory tests to verify whether the initial test is a true positive or a false positive one. Trying to explain the significance of a false positive test for AIDS or hepatitis to a healthy donor often causes fear, concern and/or anger. Thus, the use of very sensitive tests on blood donors will increase the safety of transfusion for recipients but result in loss of some donors and discard of many blood components unnecessarily. Despite the problems in applying sensitive tests to asymptomatic individuals who are not patients, the assays in place in blood banks have, nonetheless, resulted in remarkably small risks of virus transmission by transfusions. Currently, the risk of HCV infection following a transfusion is about 1 in 3,300 per unit transfused. This is an enormous improvement compared to the risks of what was called non-A, non-B hepatitis in the 1970s and 1980s before the use of the test for antibodies to HCV. For HTLV-I (and, potentially, HTLV-II) the risk of transfusion transmission is about 1 in 50,000 per unit of screened blood. Using blood which is anti-HIV-1/2 non-reactive, the risk is about 1 in 225,000 units of transmitting HIV. The risk of transfusion associated AIDS is thus quite remote in 1993. For hepatitis B virus, only about I in 200,000 units of blood transmit this virus now. In sum, only about 3 units of blood per 10,000 of those collected from acceptable, volunteer donors are currently likely to transmit a serious or fatal transfusion-transmitted viral infection. In contrast, in America, about 6 out of every 1,000 patients hospitalized will die from an accidental or preventable cause other than the underlying disease for which he/she was hospitalized. Thus, there is a 20 times greater risk that patients will die of something in the hospital unrelated to their disease on admission than of their developing a serious, potentially fatal, transfusion-transmitted viral infection from a unit of blood. Transfusions are really quite safe today.
