March 28, 2007
Approximately 2.1 million HIV tests are conducted annually in publicly funded counseling, testing, and referral (CTR) programs. However, many persons do not return for the results of conventional tests: 30% of persons who tested HIV-positive during 2000 and 39% of persons who tested HIV-negative did not return1. Almost all clients receive their rapid HIV test results because results can be provided immediately during the testing visit.
Clinical studies have demonstrated that the sensitivity2 and the specificity3 of rapid HIV tests are comparable to those of EIAs often used for screening. The negative predictive value4 of a screening test is high at the HIV prevalence observed in most U.S. testing settings (CDC, 1998). Therefore, a client with a negative rapid HIV test result can be told he or she is not infected. However, because HIV antibodies take time to develop, retesting should be recommended to persons with a recent possible exposure (sexual contact or needle sharing within 3 months). As with any screening test, the positive predictive value of a reactive rapid HIV test may be low in populations with low prevalence (see Appendix). Because some reactive test results may be false-positive, every reactive rapid test must be confirmed by a supplemental test (either Western blot or immunofluorescence assay [IFA]). (CDC, 1989).
CDC's revised Recommendations for HIV Testing of Adults, Adolescents, and Pregnant Women in Health-Care Settings, released in 2006, change some of the previous recommendations for prevention counseling for patients who take a rapid HIV test in a healthcare setting. In these settings, prevention counseling need not be conducted in conjunction with HIV diagnostic testing or HIV screening programs. This is an effort to remove a potential barrier to HIV testing in busy healthcare settings. CDC continues to support prevention counseling in all settings for persons at high risk for HIV and in non-medical settings. CDC still recommends that patients receive information about HIV testing, HIV infection, and the meaning of test results.
In addition, the 2006 Recommendations state that consent for HIV screening in healthcare settings should be incorporated into general consent for medical care rather than using a separate written consent form.
HIV counseling encompasses two components: provision of information and prevention counseling (CDC, 2001a). All clients must receive information about the rapid test and give informed consent (whether or not the consent is part of a general consent) for testing. Clients who can benefit should also receive prevention counseling.
Information can be provided either in a face-to-face meeting with a counselor or in a pamphlet, brochure, or video. Clients tested with a rapid HIV test should be given the same types of information recommended for those tested with a standard EIA:
In addition, clients tested with rapid HIV tests should be
During the initial visit, the provider can definitively tell clients whose rapid HIV test result is negative that they are not infected, unless they have had a recent (within 3 months) known or possible exposure to HIV. Retesting should be recommended for these clients because sufficient time needs to elapse in order before antibodies develop that can be detected by the test.
Further testing is always required to confirm a reactive (preliminary positive) screening test result.
Providing reactive (preliminary positive) results to clients without the benefit of a same-day confirmatory test can be a challenge. For all clients with a reactive rapid HIV test result, however, it is essential to:
A simple message to convey this information could be "Your preliminary test result is positive, but we won't know for sure if you are infected with HIV until we get the results from your confirmatory test. In the meantime, you should take precautions to avoid transmitting the virus."
Fundamentals of HIV prevention counseling with rapid HIV tests include:
With standard testing, there are always two test-associated opportunities for prevention counseling for clients who return for their results. With rapid testing, there may be either one or two.
HIV prevention counseling with a rapid HIV test completed in a single visit has been successfully implemented in numerous settings in the United States over the past several years. Experience has shown that this form of prevention counseling is feasible and is well accepted by most clients and counselors. An example of a counseling protocol for providing prevention counseling during one visit for clients receiving rapid test results and materials for counselors can be found at the RESPECT-2 web site.
Several other rapid HIV tests already being used outside the United States will likely be considered for FDA approval. Many of these tests require a single step, can be performed on whole blood, serum, plasma, oral fluid, or finger-stick blood samples, and provide results within minutes. These tests also have a high sensitivity and specificity (Branson, 2003). As these tests become available, it may be possible to implement strategies such as one recommended by the World Health Organization (WHO, 1997) whereby specific combinations of different rapid tests might be used to confirm reactive rapid HIV test results immediately.
Branson, BM. Point-of-care Rapid Tests for HIV Antibodies. Journal of Laboratory Medicine 2003;27:288-95.
Centers for Disease Control and Prevention (a). CDC. Revised Recommendations for HIV Testing of Adolescents, Adults, and Pregnant Women in Health-Care Settings. MMWR 2006; 55(RR14);1-17.
Centers for Centers for Disease Control and Prevention (b). Revised recommendations for HIV screening of pregnant women. MMWR 2001; 50(No. RR-19):59-85. PDF.
Centers for Disease Control and Prevention. Update: HIV Counseling and Testing Using Rapid Tests -- United States 1995; MMWR 1998; 47(11): 211-215.
World Health Organization. Revised Recommendations for the Selection and Use of HIV Antibody Tests. Weekly Epidemiologic Record 1997: 72(12): 81-88.
Positive predictive value is an important concept that may be difficult to understand. It depends both on the test that is used (in particular, the test's specificity) and the prevalence of infection in the population tested. An example may help to illustrate how the positive predictive value (and the proportion of false-positive test results) changes at different levels of prevalence.
We will illustrate a test that has a sensitivity of 99.9% and a specificity of 99.8%, similar to that of many rapid HIV tests and EIAs. A specificity of 99.8% means that 0.2% (2 tests out of 1,000) will be false-positive. For this example, we will test 1,000 persons, first in an STD clinic with high HIV prevalence: 5%. Testing 1,000 persons, we would discover 50 persons who were truly positive. Based on the test's specificity, we would also encounter 2 false-positive test results. Thus, the positive predictive value of a reactive test in this setting would be (50 true positive tests) divided by (52 total positive tests) or 96%.
Using this same test in a population with low prevalence gives us a very different predictive value. For this example, we will use the same test in a family planning clinic, where the HIV prevalence is 0.1%. Testing 1,000 persons in this clinic, 1 person would be truly positive, but again, 2 test results would be false-positive. The positive predictive value of a reactive test in this setting, therefore, would be (1 true positive test) divided by (3 total positive tests) or 33%. Notice that in both these examples, the number of false-positive tests is the same, but the proportion of false-positive tests is very different.
The following table shows the positive predictive values at different levels of HIV prevalence for a test with 99.8% specificity.
Positive Predictive Value of HIV Tests in Populations with Differing HIV Prevalence
|Positive Predictive Value|