Can we keep our blood supply safe from AIDS?
Since 1982, when the first transfusion-transmitted case of acquired immunodeficiency syndrome was reported, blood bankers have been put on the defensive to assure the medical profession and the public that the blood supply is safe from AIDS. In this article, we will explore whether a conclusion that our blood supply is safe is consistent with available data.
Without addressing the question of "how safe is safe," we will look at the two measures undertaken to eliminate contamination of the nation's blood supply by human immunodeficiency virus (HIV, originally called HTLV-III): self-deferral of high-risk donors and the ELISA test to detect HIV antibody in collected units. We will consider how successful these measures have been and what further steps, if any, are necessary.
The effectiveness of self-deferal can be measured in several ways: 1) we can look at the difference--or gap--between the prevalence of HIV among donors and the prevalence of HIV in the non-donor community with similar demographics; 2) we can measure the decrease in the percentage of HIV-infected donors who have donated units of blood since the onset of the self-deferral program; and 3) we can examine whether the prevalence of confirmed HIV seropositivity among donors segregates independently from the prevalence rate of the community itself. We will inspect each of these in turn:
1. Determining the HIV-prevalence gap between donors and the community at large is not as easy as it first seems because donors are not a microcosm of the entire population. They represent a select subset with specific demographics. To determine whether a prevalence gap exists, it would be necessary to develop a set of demographically matched controls who do not serve as blood donors and whose risk factors for HIV are identical to those of the community at large. To my knowledge, this has never been undertaken.
2. As for achieving a decrease in the percentage of infected donors, the American Red Cross has data indicating self-deferral is quite effective. When the HIV testing program began in April 1985, there were 3.8 Western blot-positive donors per 10,000 donors. Recent data reveal a significant drop to 1 per 10,000. However, as encouraging as these data are, there are indications they may be misleading.
3. What about the final yardstick for the effectiveness of self-deferral? At the San Francisco Irwin Memorial Blood Center, the prevalence of Western blot-confirmed positives is 10 per 10,000 donors, 10-fold greater than that of Red Cross donors. Admittedly, San Francisco has a remarkably high prevalence of AIDS. But if self-deferral really worked, the prevalence of seropositive donors in San Francisco would be no higher than the 1 per 10,000 observed elsewhere in the country by Red Cross. In other words, the prevalence of seropositive donors should segregate independently from the prevalence of HIV within the community from which they come. The fact that a 10-fold difference in seropositivity exists invalidates the hypothesis that self-deferral programs are an adequate first line of defense against transfusion-transmitted HIV infection.
There are three reasons why self-deferral programs may not operate effectively. The first is that donors may come to the blood center chiefly to get free HIV antibody tests. The second is that the donors may not be aware they are in a high-risk group, and the third is that they may be aware, but either do not care or choose to deny their high-risk status.
The first explanation is unlikely in San Francisco at this time. Anxious individuals can obtain HIV antibody tests at many alternate test site facilities within the community. In addition, the built-in three- to six-month interval between testing and notification of seropositive donors has been designed to keep such individuals from using blood centers in this dangerous manner.
On the other hand, there is considerable evidence that the other two factors are weakening self-deferral. At a National Institutes of Health consensus development panel convened in July 1986, several blood bankers discussed their experiences. Harvey Alter of NIH reported that 25 per cent of a small group of homosexual/bisexual men did not consider themselves to be in a risk group, and another 25 per cent did not understand or had not read the literature presented to them at the blood bank. Steven Kleinman of the Los Angeles Red Cross said many bisexuals did not perceive a risk of HIV infection. Thomas Zuck, director of the Food and Drug Administration's division of blood and blood products, stated that "experience during the first year of testing also suggests that people who are at increased risk of HTLV-III [HIV] infection continue to donate blood and plasma."
Typical of this phenomenon is the case of a 20-year-old seropositive woman who was the sexual partner of a 25-year-old hemophiliac. She had read a voluntary blood donor deferral brochure, which identified sexual partners of those at risk as persons who should not donate--nevertheless, she did not believe that she was at risk! A recent survey of members of a "swingers" club in Minnesota found that 73 per cent of these high-risk heterosexuals did not consider themselves to be at risk.
Let us turn our attention to the ELISA HIV antibody test, the second program designed to assure the safety of the blood supply from AIDS. Many blood bankers cite the 99+ per cent sensitivity of the test as evidence that it is highly effective, in conjunction with voluntary deferral, as a means of protecting the blood supply.
Take a closer look at this hypothesis. First of all, it is hard to understand how one determines the sensitivty of the HIV antibody test without applying Koch's postulates. At present, no gold standard can be applied. For example, do we count, as part of our denominator in calculating test sensitivity, infected individuals who are in the early phase of their illness, in whom antibody has not yet developed? What about individuals with terminal AIDS who have lost their antibody as a result of disease? Are they included?
Cultures are not helpful as a gold standard, because virus is recovered from infected individuals only about 60 per cent of the time. (Some researchers report a higher percentage of positives.) And finally, there have been multiple reports of culture-positive, antibody-negative individuals who are not in the initial six-week seronegative window. So the concept of sensitivity needs further clarification--or at least several caveats.
Even if we accept the stated 99+ per cent sensitivity of the antibody test using the classic understanding of the term--i.e., the per cent of individuals with the condition who test positive--it is apparent that a few HIV-positive units will slip through the cracks as a result of less than 100 per cent sensitivity. These units will result in transmission to others as a function of the number of blood products prepared from each donation and the infectivity of a single unit of contaminated blood, which is about 90 per cent. HIV-positive units will also get through because of clerical errors and, most important, because of the prevalence of infection in the population under consideration.
Jerry Kolins of the Community Blood Bank of North County Escondido and Poway (Calif.) recently calculated that about 24 infectious units per year would be transfused, while Harvey Sapolsky at Massachusetts Institute of Technology estimated that 18 to 25 cases of transfusion-transmitted AIDS would occur annually. As shown below, my calculations suggest these numbers are too conservative.
Consider the implication of the prevalence of HIV infection as it relates to false-negative antibody results, and the impact this may have on the safety of transfusion now and in the future as HIV prevalence increases. At present, the Centers for Disease Control estimates between 1 million and 1.5 million infections in the U.S. It also estimates that the number of new infections continues to double about every year. If we take the average, 1.25 million, and assume that's how many additional infections there will be during the next 365 days, that means transmission of 3,425 new infections of HIV each day.
Available data suggest that antibody conversion occurs about six weeks (42 days) after infection has been acquired. This means that at the present time there are 3,425 X 42 or 143,850 HIV-infected individuals who will test negative for antibody. Presumably the majority of them will be viremic and therefore high-risk donors, but no data are yet available as to how many days must pass before a newly infected individual can transmit the infection.
For the purpose of illustration, lett us suppose that there are no policies in effect to encourage self-deferral, and also that the rate of voluntary donation among this infected-seronegative group is identical to that of the population as a whole. Since about 4 per cent of eligible donors supply all of the blood needs of the country, 5,754 (.04 X 143,850) of these infected individuals would donate each year. The HIV antibody test would fail to detect every one of them.
Although the average donor donates 1.46 times a year, for our purposes we will assume that by the second donation, seroconversion already will have occurred. Our calculations are based on the premise that only one unit per infected donor is capable of evading detection by antibody testing.
The average unit results in up to 2.07 transfusable products. (This represents current data in Clark County, Nev., and does not include plasma that is obtained for fractionation purposes.) Multiplying that figure by the number of infected-seronegative donors (2.07 X 5,754) yields 11,911 infected blood products. The percentage of units that outdate is no greater than 5 per cent, so at least 95 per cent or 11,315 of these products will be transfused. Since 90 per cent of individuals transfused with HIV-tainted blood seroconvert, we would anticipate 10,183 transfusion-related HIV infections per year (see Table I).
In the final analysis, the success of self-deferral can be measured by its record against these 10,183 infections. If 50 per cent of the infected-seronegative donors were to stay away, 5,092 cases of transfusion-transmitted HIV would occur over the next 12 months. Even if we discourage 99 per cent from donating, there would still be 102 new cases.
Note that, with the probable exception of gay males, the others at risk for HIV infection--intravenous drug users, prostitutes, hemophiliacs, and Haitians--are probably underrepresented as volunteer blood donors. So the 4 per cent figure in my calculations may be artificially high; in that case, the annual number of cases of transfusion-transmitted HIV is overestimated.
The extent to which transfusion-transmitted HIV is perceived as a problem with the safety of the nation's blood supply depends on several variables: the actual number of infections transmitted, the percentage of transfusion recipients who survive long enough after transfusion to seroconvert and to develop AIDS-related complex or AIDS, the percentage of transfusion recipients who are tested several months following transfusion to determine whether they were indeed infected, and the attention that these cases receive from the medical and lay press.
The percentage of patients surviving depends upon the reasons they were transfused. Among oncology patients, the long-term survival rate is low, while among trauma patients who get past the initial resuscitation process, it is quite high. In one study, it was estimated that 60 per cent of those who received infected units would die from the illness that required the transfusion. In addition, since ARC or AIDS develops over a period of years, many infected recipients will not survive long enough to develop immunodeficiency--or in many instances, even long enough to seroconvert. More data are needed to develop projections about patient mix and life expectancy following transfusion in relation to seroconversion, ARC, and AIDS.
In conclusion, it is apparent that as the HIV epidemic spreads, the problem of tainted blood will be a greater cause for public concern, despite our temporary success at decreasing the transfusion of HIV-infected products through self-deferral and antibody screening. The extent to which self-deferral can be perfected, coupled with our ability to develop new lab tests that will identify the already large--and rapidly growing--pool of infected seronegative individuals, will determine how safe our blood will be.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||not likely, even in view of current testing|
|Author:||Soloway, Henry B.|
|Publication:||Medical Laboratory Observer|
|Date:||Jan 1, 1987|
|Previous Article:||Quality control in the new environment: QC materials; an ideal control closely simulates patient specimens, is stable, and comes in large homogeneous...|
|Next Article:||In-house software monitors and projects costs.|