TMU Title Bar

April, 1999



Leo Manack, M.D., Fellow in Transfusion Medicine

Joseph E. Kiss, M.D., Medical Director, Hemapheresis and Blood Services

Recent media coverage of Hepatitis C Virus Lookback and Central Blood Bank’s health advisory about a potential risk to recipients of blood transfusions has once again focused attention on the important issue of transfusion safety. Since we last addressed this issue in 1995, further gains have been made to ensure the safety of the volunteer blood supply. Increasing vigilance in donor screening, coupled with advances in viral detection continue to make the blood supply the safest it has ever been.

The Importance of Donor Screening

Donor screening is crucial because it defers those who have risk factors for transmissible diseases and may be donating during the period between infection and seroconversion. During this window of time, the donor’s blood may be infectious, but lacking the antibodies or antigens detected by most screening tests. The intervals for the four transfusion transmitted viruses that we routinely test are: HIV, 16 days; HTLV, 51 days; HCV, 82 days; and HBV, 59 days.

Donations during this so called "window period" represent the greatest risk to the blood supply, and consequently, blood bankers have focused on efforts to shorten this interval.

Closing the Window

The push to shorten the window period began with the introduction of HIV p24 antigen testing of donor blood in 1996. The p24 antigen is present approximately 6 days earlier than its antibody. An estimated eight "window period" infected donations per year were expected to be interdicted by this testing nationwide. However, in practice, there appears to be fewer antibody negative/p24 antigen positive donations than once thought, apparently because those who are recently infected with HIV develop mononucleosis-like symptoms, which would exclude them from donation. The additional cost of

p24 antigen testing (between 120 and 160 million dollars annually) may not seem cost effective to some. However, in 1998 one Central Blood Bank donor was found to be p24 antigen positive with a negative HIV antibody test. In this instance, HIV infection was prevented in at least two and possibly three transfusion recipients (based on component fractionation of the donor unit into packed RBCs, platelets, and plasma).

Blood collection agencies in the United States will soon go even further and initiate direct viral nucleic acid testing of donated blood products. This new screening method will be discussed in detail in the next TMU. The increased sensitivity of the testing procedure will reduce the window period for HCV to 10-30 days and HIV to 10 days, resulting in a marked improvement ( 5-fold lower) in the safety profile for HCV, and to a lesser extent for HIV.

Assessing the Risk

Probably the best way to calculate the risk for infection associated with transfusion is to prospectively evaluate recipients for the development of infection. The current very low risk of transfusion-transmitted infection makes such studies extremely difficult to perform. These studies would require enormous numbers of recipients to detect any risk.

Again, the model of "window period" donations is useful for estimating the risk associated with transfusion. This model rests on the assumption that screening tests are extremely sensitive, that the antibodies detected by those tests are reliably present after the window period has elapsed, and that there is an equal probability of infection throughout time. The risk per unit transfused is computed as the product of the incidence of infection in the donor population and the proportion of time a donor may be in the "window period" before seroconversion1:


RISK = incidence x window period in days

Based on 1997-98 incidence rates in Central Blood Bank repeat donors, the estimated risks of transfusion-transmitted viral diseases from a single unit of blood are as follows:




1 / 1.25 million


1 / 135,000


1 / 133,000


1 / 370,000


It should be pointed out that the confidence intervals (ranges) associated with the above calculations are large and they do not take into account viral incidence rates in first time donors, which may be higher than repeat donors. However, the percentage of first time donors is relatively small (<20%) and other factors may reduce the risk further. For example, some contaminated units would be excluded from use because of abnormalities in screening tests that function as surrogate markers for viral infection. In addition, we are assuming that each contaminated unit will indeed transmit virus, when in fact, for a virus such as HTLV, the likelihood of seroconversion after receiving blood from an infected donor is only 1 in 4.

Less Risk Here

How does the safety of the blood supply in the tri-state region compare with that of the nation? Considering our relatively stable and older donor population, and the low prevalence of viral infection in the area, quite well indeed. The risk for HIV and hepatitis infection in our area is approximately one-half less than published national estimates.

Perception At Odds With Reality

The above discussion presupposes that viral testing is performed under optimal circumstances, using current "good manufacturing practices" (cGMPs). Because of concerns raised in regard to the adequacy of testing performed for Central Blood Bank between the years 1991-1996, the blood center issued "Public Health Announcements" offering testing to patients who received blood transfusions in area hospitals. Between November 1998 and March 1999, over 6,000 patients came forward to be tested. Only a small number - 70 (1.2%) - had positive test results, mostly for hepatitis C. (It is important to note that of this number, 58 (1%) were HCV positive, significantly less than the Centers For Disease Control and Prevention (CDC) estimate of 1.8% in the general population.) Ongoing investigations have found no evidence to date that any of the positive tests arose from improper testing of the blood products the patients received, with most individuals having preexisting infection or other risk factors for infection. While the results reaffirm the low risk of transfusion-transmitted disease, one lesson is clear from the sometimes emotional reaction in response to the public health advisory: public expectations are high when it comes to blood safety.

Putting Risk into Perspective

Some perils are thought to be far more pervasive than they really are. It seems that the rare transfusion associated viral infection, like the uncommon airplane accident, has exaggerated the concern over such a mishap. However, fear of flying can overshadow the danger in crossing the street. So too, with medical care. There are discrete risks - viral and non-viral- associated with blood transfusion, whether allogeneic or autologous. However, the risk of fatal anaphylaxis from a single dose of penicillin is at least 2-fold greater, while the risk of death as a result of surgery is 30-fold greater. Indeed, it has been suggested that a blood transfusion may be one of the safest invasive medical procedures performed in hospitals today.


1. Schreiber GB, Busch MP, et al. The risk of transfusion transmitted viral infections. NEJM 1996; 334:1685-92.


Copyright 1999, Institute For Transfusion Medicine

For questions regarding the Update On The Risk of Viral Infection From A Blood Transfusion In The Tri-State Region, please contact Joseph E. Kiss, M.D. at: (412) 209-7326

Copies of the Transfusion Medicine Update can be obtained by contacting Deborah Small (412) 209-7320

Visit our web page for information and previous issues of the TMU newsletter at