6.3 Screening strategies

6.3.1 General

6.3.2 Infectious agents screened for

6.3.3 Number of donations screened (selection of assay format)

6.3.4 National and regional health policies

6.3.5 Specific screening strategies

    The strategies selected for donation screening have to reflect the national screening policy and are basically determined by the infectious agents screened for, the number of donations screened and national policies. For any situation there will always be alternative strategies, all of which may have good as well as bad features. No one strategy is likely to be perfect, but recognizing this fact is important in assessing the relative merits of any strategy. However, choice of the most appropriate strategy is vital to ensuring that the screening programme is effective.

6.3.2 Infectious agents screened for

    There are a number of issues to be considered when deciding which infectious agents are to be screened for. Although the national screening policy should indicate which agents are to be screened for, for each infectious agent the most appropriate circulating marker of infection needs to be identified. In some cases this is not too much of a problem as the number of markers detectable is limited and it is quite clear which marker should be used for each infectious agent. In other cases, however, this may not be as clear and careful selection of the most appropriate marker is needed. It is also possible that the situation for any infectious agent may change after the screening strategy has been designed and implemented; all strategies must be monitored and changes made when needed. The choice of which marker to screen for needs to reflect the infectivity of an infected donor at any stage of infection, the circulating markers of infection produced by the infectious agent and the predictive value of each marker.

a) Infectivity

    Most infectious agents give rise to particular serological patterns of infection, but these can be generalized into four stages: incubation, acute infection, recovery, immunity. In some cases chronic infection ensues, and recovery and acquiring immunity may take years. In other cases infection persists for the lifetime of the individual. However, taking the general four-stage model, the strategy first needs to understand at what point screening is going to identify an infectious donor.

    As examples, let us look at the infection profiles of hepatitis B virus and HIV (see Chapter 3 for detailed information on the viruses). If we first look at acute hepatitis B virus infection, here we can see a relatively simple picture. Following incubation the first marker of infection to appear in the bloodstream is HBsAg. This is produced in very large quantities during acute infection and marks a highly infectious stage of infection. The production of antibodies to viral proteins follows the appearance of HBsAg and ultimately marks resolution of infection and development of immunity. HBsAg assays are widely available and relatively cheap, and have good sensitivity and specificity.

    If we now look at HIV infection, the picture is quite different. Following incubation the first marker of infection to appear in the bloodstream is pro-viral RNA, 5_10 days after this HIV p24 antigen and HIV DNA appear, followed 5_10 days later by anti-HIV which persists, probably for the lifetime of the individual. In HIV infection anti-HIV does not represent immunity; rather it demonstrates previous infection and probable persistence of the virus in the bloodstream. Although HIV RNA, p24 antigen and HIV DNA are all produced before anti-HIV, their levels in the bloodstream fluctuate so that screening for only one of them may result in the failure to detect an infected individual, while anti-HIV, once produced, persists, and at a level that is easily detectable. Indeed p24 antigen levels fall below detectable levels very quickly after anti-HIV appears. In addition, assays for RNA and DNA are currently more specialized and not suited to mass screening procedures; p24 antigen assays are available in an EIA format but are expensive and need to be used in conjunction with anti-HIV assays. Thus anti-HIV is the single best marker of HIV infection for the screening of donated blood collected from a population with a low to medium rate of acquisition of HIV.

    As can be seen, the marker of infection screened for is dependent on the particular infectious agent and may represent a compromise between as early detection of infection as possible and a practical and workable screening programme. This has to be reflected in the screening strategy selected, but may be affected by the prevalence of the infectious agent in the donor population. Again, HIV is a good example of an agent for which the screening strategy may alter depending upon the prevalence in the donor population (see Chapter 3 for a more detailed discussion on HIV antigen testing). As the prevalence increases it has been argued that the possibility of detecting an HIV-antigen-positive but anti-HIV-negative donation increases and adding HIV antigen testing to the HIV testing strategy may have to be considered.

b) Use of surrogate markers

    Screening for surrogate markers has been performed in some countries for many years, while other countries have never used that strategy. Surrogate markers are nonviral specific markers that are thought to provide indirect evidence of viral infection where no specific markers can be found. In some countries serum liver aminotransferase levels, specifically ALT, were used to try to reduce cases of PTNANB hepatitis before anti-HCV assays were available, but are now used to try to prevent cases of HBV transmission from HBsAg-negative individuals and transmission of other hepatic viruses. Generally the introduction of anti-HCV screening has demonstrated how poorly predictive ALT screening is, and a number of countries have ceased or are actively considering ceasing ALT screening. In a similar way, anti-HBc screening was also used before anti-HCV screening assays became available to try to reduce cases of PTNANBH, but this screening now has no value in this context.

6.3.3 Number of donations screened (selection of assay format)

    Once the particular agents to be screened for and the individual markers of infection have been decided, the number of donations screened is probably the next major factor to consider. This is important because it essentially determines the most appropriate format of screening assay to use; the major factor in deciding on a particular format is the number of tests performed in a single test run and the frequency of test runs. As previously discussed there are currently three basic formats of screening assay that are commercially widely available: simple/rapid discrete assays, particle agglutination assays (PAs) and enzyme immunoassays (EIAs). There are a number of different approaches that can be taken in relating the number of tests to assay format, but perhaps the simplest is to link the number of donations tested per day or week to assay format. The following schema is a simple and workable approach which could act as a basis for developing more precise individual schemes.

• When testing 1-35 donations per week use a rapid/simple assay.

• When testing 35-60 donations per week use a particle agglutination assay.

• When testing more than 60 donations per week use an EIA.

    Such an approach is not only an effective use of resources, but also ensures that the availability of blood is compatible with local needs. On the one hand, a small transfusion laboratory does not need the automated equipment required to perform EIAs when the workload is small and there are few staff; such equipment would be totally inappropriate and would probably rarely be used. In this situation a supply of rapid/simple tests would be quite adequate to ensure screening prior to use of the blood. In addition, in such a situation it is highly likely that blood is either collected only when needed, or that what is collected is used very quickly; in either case stocks of blood are rarely held. However, using such rapid tests, freshly collected blood can be tested and made available immediately. On the other hand, a large laboratory would have the need and resources to utilize the automated equipment needed to perform EIAs. Because EIAs are slower and more involved procedures, it takes longer to test the blood, but much larger numbers can be tested at any one time. Such a laboratory is, however, more likely to have a more developed blood collection programme and to hold stocks of blood so that blood needs can be supplied from existing tested stock without needing emergency collection and testing.

    All three main assay formats have their own advantages and disadvantages, and these can obviously make one format particularly attractive and more appropriate in a certain situation. Table 6.1 lists some of the more important advantages and disadvantages associated with the different assay formats.

6.3.4 National and regional health policies

    The national blood screening policy should be part of the general national health policies. However, there may be other policies, national or regional, that affect the screening strategies developed either at national or a more local level. These commonly involve donor rather than donation issues, often to do with the confirmation and notification of positive donors, and may restrict the transfusion service in the management of positive donors.

• It may be policy not to inform donors of a confirmed positive screening result, but only to discard the blood and not recall the donor. Such a policy is uncommon but may reflect the fact that donor identity can often not be confirmed (a donor may give a false name and address). It may also reflect the inability of the health care system to provide appropriate care for such an individual, who may appear to be healthy and well. In such situations the supply of blood is often critical to the patient's survival, and some blood safety issues simply cannot be addressed properly. As long as screening is performed before transfusion and the screening programme is reasonably well designed and controlled, any risk of infection is at least minimized and such a policy can be largely justified.

TABLE 6.1 Advantages and disadvantages of different assay formats

• It may be policy not to confirm the screening results, but to inform the donor on the basis of the screening result, either a single result or on the basis of repeat testing. There are a number of reasons why such a policy may be adopted, the main one being lack of confirmatory facilities, but also there may not be a dedicated reference laboratory and the different assays needed to reliably confirm screening results may not be available. It is clear that reliance upon a single screening result to exclude donors is very wasteful of blood as many assays initially give false-positive results, many of which disappear on repeat testing. However, either time or cost constraints can limit the ability to repeat any initially reactive donations, making the use of the initial result essentially unavoidable. A repeatably positive result has a better predictive value than a single result, depending upon the specificity of the test and the prevalence of the particular agent in the population, and counselling on this basis can be more appropriate.

• It may be policy not to reinstate donors who are reactive on screening but confirmed negative. For many years it is been recognized that some donors will give false-positive results with some assays. When confirmatory testing is performed these donors are shown to be truly negative, just falsely reactive in the primary screening test used. In addition, in many of these donors this reactivity is transient, often only affecting one or two donations; thus, these donors may be deemed only temporarily unsuitable. Some national policies, however, may permanently defer such donors on the basis that any repeatably reactive screening result should be considered relevant. This is very wasteful of donors and is not really an appropriate policy with the range of assays now available to confirm the initial screening results and the wealth of knowledge available about such donors.

6.3.5 Specific screening strategies

    Taking into account all of the previous comments and discussion, some specific screening algorithms are now examined and their specific applications discussed.

a) WHO strategies

    In 1992 WHO published some basic screening strategies that could be applied to HIV testing in a number of different circumstances depending upon the reason for testing (WHO Weekly epidemiological record, 1992, 20:145149). Three basic strategies were presented and suggestions made about the suitability of these strategies for different testing scenarios. Table 6.2 suggests possible uses of the testing strategies outlined below.

    Strategy I. Samples tested with one EIA or rapid/simple assay. Any reactive sera considered to be anti-HIV positive.

    Strategy II. Samples tested with one EIA or rapid/simple assay. Reactive sera retested with second assay but based upon different antigen of test principle. Sera reactive in both tests considered to be anti-HIV positive, sera reactive in only first test considered to be anti-HIV negative.

TABLE 6.2 HIV testing according to strategy

Strategy III. As strategy II but a third test is used for sera reactive in the first two tests. Sera reactive in all three tests considered to be anti-HIV positive, sera reactive in only the first two tests are considered to be anti-HIV indeterminate.

    Although strategies I to III can be made to work in blood screening, they were designed on a more universal basis to provide common strategies to cover different screening rationales. There are other slightly different approaches that are more specific to blood screening and commonly used by many transfusion services.

Strategy 1 Use WHO strategies I/II. Discard all reactive donations.

Strategy 2 Repeat any initially reactive samples in duplicate using the same test. Discard any donations reactive in either of the repeat tests.

Strategy 3 Repeat any initially reactive samples in duplicate using an alternative test of equal sensitivity and the same principle. Discard any donations reactive in either of the repeat tests.

    Of these, strategy 2 is the one most commonly used in developed countries. Strategy 3 is increasing in acceptability; it is most useful if the specificity of the initial assay is not optimal; and a number of donations would be repeatably reactive, although subsequently confirmed negative, if strategy 2 were to be used, and discarded. Use of strategy 3 would reduce significantly wastage of blood.

b) Pooling

    All of the six strategies discussed in item a) above are based on the testing of single discrete samples. The pooling of samples prior to testing is a very controversial issue. Despite the argument of cost, the disadvantages of pooling include potential loss of sensitivity and the requirement of individual laboratory validation. Additionally if the prevalence of the infective agent in the donor population is too high the number of positive pools becomes too high and the amount of additional testing needed to identify and confirm the positive samples exceeds the savings in cost that would be gained.

    The health of an individual is the highest priority, and all optimal measures must be taken to minimize the risk of a transfer of a disease. It is a common rule that a specimen from an individual received by a laboratory for diagnostic testing for diagnosis or exclusion of a disease should never be pooled with another specimen. It would be against ethical and legal principles to mix specimens for testing knowing that there is a potential loss of sensitivity. The courts of justice in some countries in Europe (e.g. in France and Germany) have condemned several laboratories because they deliberately ignored this rule. The argument of those who pool, that it is better to pool the samples and accept the weaknesses rather than not to screen at all, does not justify recommending pooling. Policy-makers may take it as a standard and acceptable rule in all situations that pooling should not be practised. The following are among the technical arguments against pooling:

1. Each method (even the most sensitive one) has a limit of detection. The limit of detection is the lowest concentration at which an agent can be positively identified. The limit of detection of a method is determined by diluting the concentration of an agent to be measured. Pooling a specimen with other specimens in clinical routine or for blood transfusion will automatically result in the same effect. This means that an agent that may be present in a specimen at a concentration near to the limit of detection of a method, will not be detected by the test when the sample is diluted to a concentration below this limit. This also means that diluting a specimen might create a condition of measurement that is less than optimal.

2. Each test system that is used by a clinical laboratory is more or less specific. Enzyme immunoassays (such as for HIV antibodies and hepatitis B surface antigen) measure an agent by means of an immunological reaction. Such a reaction can be interfered with by non-specific interferents that might be added when pooling blood or a serum specimen with other blood or serum specimens. For example, certain immunoglobulins (such as rheumatic factors) can interfere with an immunological test. The interfering substance can therefore inhibit the test for diagnosis. A negative test result caused by such interference is not caused by a dilution and may occur at a concentration of an agent where one would normally expect a positive test result.

3. Testing blood for transfusion as well as for the preparation of blood products for therapy of patients requires the highest possible standard, and optimal conditions of investigation must be chosen to exclude even a minimal risk of infection. This is why the WHO Expert Committee on Biological Standardization (ECBS) has recommended that each donation be individually tested for infective agents. Moreover, when blood donations are pooled for the preparation of blood products, the pools must be tested during the manufacturing process in addition to the first testing of each individual donation.

    Pooling might be acceptable for epidemiological studies of population.