In most developing countries the clinical diagnosis of HIV infection is made after serological testing, where an indirect diagnostic method identifies the presence of HIV antibodies in an individual's serum. The most common method for detecting antibodies is the Enzyme Linked Immunosorbent Assay (ELISA) test, which requires a machine to measure color change in test wells. Because antibodies to HIV develop weeks after the initial infection, antibody tests will be negative in the "window period" between infection and an antibody response. Newer, "rapid" tests rely on agglutination/absorption methods and color change, which is visible to the naked eye. In some circumstances (for instance, in a child born to an HIV-positive mother), more direct diagnostic methods are available to detect the virus and not the immune response; however, these are rarely used outside industrialized countries. Three common tests for direct viral detection are P24 antigen detection (P24 is a protein on the HIV virus); isolation of the virus by culture; and amplification of the HIV genetic information (polymerase chain reaction or PCR).
The World Health Organization (WHO) recommends three testing strategies, depending on the HIV prevalence in the population, on the sensitivity and specificity of the antibody tests being used, and on whether the objective is blood screening, surveillance or diagnosis. Because HIV testing in voluntary counseling and testing (VCT) centers is performed for diagnostic purposes, it must be highly sensitive (detecting all the true positives) and highly specific (no false positive tests). Repeat testing of positive specimens with different test formats increases the specificity of the result.
Today there are dozens of high-quality HIV testing kits on the market. Testing protocols that maximize both the sensitivity and specificity of antibody detection must be incorporated into VCT services. It is normally the responsibility of government regulatory bodies — Ministries of Health, National AIDS Control Programs, National Reference Laboratories — to develop the most feasible testing strategies for the country. Factors to be considered include:
- scientific validity;
- existing laboratory infrastructure;
- the presence and capabilities of a reference laboratory for quality control;
- volume of testing required (measured by number of people to be tested daily)
- clients' preferences; and
- impact of protocol on delivery of health services (same day results versus return appointment)
- long-term costs.
The wide availability of high-quality rapid HIV tests has helped VCT programs overcome one of their main problems — a failure of patients to return to obtain test results. Using ELISA-based technology, VCT clients must typically wait at least 72 hours for their results. Experience in most VCT centers suggests that 20 percent to 40 percent of clients tested for HIV do not return for their test results. In settings where service utilization has been low, introducing rapid tests has increased access to VCT services significantly. Rapid HIV assays have recently been shown to have important applications for HIV prevention as well as for care and support intervention. For those who are negative, results can be communicated on the same day, providing an additional avenue for prevention strategies. Those who are positive can be referred early to clinical care and support services. Some view the waiting period for test results using ELISA technology as a time for reflection, giving pre-test counseling a greater impact. This can be accomplished with rapid testing by ensuring time between the pre-test and blood-draw or finger-prick stage. This rapid turn around of HIV test results has a significant impact on the design of interventions to prevent mother-to-child transmission, on preventive therapy for opportunistic infections and on post exposure prophylaxis.
The advent of the rapid test technologies, especially those that use whole blood instead of serum, has also increased the options for testing protocols and procedures to accommodate different service settings and client preferences. These options are serial testing and parallel testing.
-
Serial testing: In serial testing, all persons are tested with a rapid HIV test. If the test is positive, a second, different rapid HIV test is performed. Discordant test results are further tested with a third type of rapid HIV test. The tests are performed in series.
-
Parallel testing: In parallel testing, all persons are tested using two tests simultaneously (in parallel). If the tests are discordant — which is estimated to occur less than one percent of the time — a third type of rapid test is used as "tiebreaker."
Advantages and disadvantages to these approaches are outlined below:
Serial testing
-
This is the protocol recommended by both WHO (in 1997) and the Centers for Disease Control and Prevention (in 1998). But these recommendations were based on test systems that used serum or plasma, not whole blood.
-
Several published studies have demonstrated highly accurate test results (predictive values) comparable to ELISA's in sensitivity and specificity. The results are equivalent — and in some ways better than — the standard algorithm of ELISA followed by Western blot. These studies also conclude that this approach is accurate and cost-effective.
-
Drawing of a veni-puncture sample of blood (one option) makes it possible to have additional sera for use in a second HIV test, if necessary, or to test for other relevant diseases, such as syphilis or hepatitis. The extra sample can be archived for quality assurance. If a finger-prick sample were used, the client would have to be called back to give a sample for a second test.
- This process requires a longer waiting time for clients who test positive on the first test. Good patient flow management could address the issue of longer waiting times. This waiting time might be used to provide health information and to minimize anxiety. Video or other distractions may be provided.
- This approach is less expensive than parallel testing given the current cost of HIV testing kits, blood tubes and needles. Savings depends on the HIV prevalence in the client bases (higher prevalence results in a greater number of second tests in the serial testing system). But at least half of the tests used in the parallel system would also be administered in a serial system.
Parallel testing
- Several published studies find no significant difference in accuracy between parallel testing and serial testing.
- Clients perceive that two tests are better than one. This increased trust in the VCT center reduces "shopping around" and helps win public trust in the services.
- Having two tests done with only one finger-stick reduces the potential for stigma that might result if patients are called back for a second visit, which would be the case for serial testing systems using rapid HIV tests.
-
Shorter clinic waiting times resulting from simultaneous testing might reduce the time taken from work, benefiting the client. The shorter waiting time might also lessen the anxiety experienced while waiting for results.
-
Blood samples taken by finger prick are more suitable for field conditions (mobile clinics, remote areas, doctor's offices) than is veni-puncture.
Resources
- WHO (Weekly Epidemiological Record #12, Mar 1997).
- CDC (Mar 27, 1998 Morbidity and Mortality Weekly Report).
- Wilkinson et al. AIDS 1997,11:337-381.
- Anderson et al. AIDS 1997; 11; 1815-1822.
- Kobalvi-Dème et la. Journal of Clinical Microbiology; Vol. 39 No5 May 2001, p 1808-1812.
