Introduction

Testing for HIV infection is useful to identify HIV-infected persons who may benefit from early medical intervention, to identify HIV-negative persons who may benefit from risk reduction counseling, for epidemiologic monitoring, and for public health planning. Concerns about confidentiality, public and occupational health and safety, civil liberties, and ethics are raised by each available testing methodology and program. This chapter is a review of the background and proposed indications for HIV testing.

Technical aspects of specific tests, including standard and rapid detection of anti-HIV antibodies in blood and saliva, detection of HIV RNA by the polymerase chain reaction (PCR) and branched chain oligonucleotide (bDNA) assays, and culturing of live HIV are discussed in other chapters. Identification of people with acute primary HIV infection also is not included in this discussion. Testing of health care workers as well as testing after a high risk sexual or injection drug use exposure to HIV with the purpose of offering post-exposure prophylaxis to those who are HIV antibody negative are discussed briefly in this chapter and in other chapters.

HIV testing may be indicated for several reasons:

Each of these circumstances is discussed after considering common issues.

General Considerations

Testing serum for antibodies to HIV with a standard ELISA (followed by a confirmatory Western Blot) is currently the most common, cost-effective, and accurate method of screening for infection.(1-5) Rapid serum HIV antibody tests, saliva- and urine- based antibody tests, and home HIV antibody testing kits have been approved by the Food and Drug Administration (FDA) and are being marketed.(6-8) HIV RNA tests are being used in research and clinical settings to diagnose primary HIV infection before the formation of detectable antibodies.

Antibody testing has revealed critical information about the epidemiology of HIV infection and the mechanisms of HIV transmission. Viral load testing has helped to define the natural history and prognosis of cohorts of people with HIV infection. There is strong and growing evidence that asymptomatic and minimally symptomatic HIV-infected persons benefit from early diagnosis and treatment.(9-13) Decisions about medical care, pregnancy, sexual and injection drug use behaviors, and career planning may be affected by the results of an HIV antibody test. Thus, there is a strong rationale for voluntary testing based on an understanding of the risks and benefits involved and informed consent.(9, 14) No strong rationale exists, however, for a policy of mandatory testing - testing people without their consent - for the purpose of controlling the HIV epidemic.

The risk that HIV test results may become known must be recognized by testing counselors, clinicians, public health policy makers, and, above all, persons who are considering being tested. Unfortunately, persons with HIV confront not only fear and pain from knowing they are infected, but also serious social, financial, and emotional problems resulting from individual and institutional prejudices about people with HIV and AIDS.

Eviction, job loss, inability to buy or maintain health or life insurance, and abandonment by family and friends are examples of some of the unfortunate consequences of being labeled as infected with "the AIDS virus." These events are not rare and can occur whenever confidentiality is violated or test results are requested by and released to employers, insurance companies, or other individuals and groups.

Antibody tests occasionally may be falsely positive, even when carried out and interpreted properly.(15) Influenza immunization may temporarily cause a false-positive antibody test.(16) Erroneous test results and improper interpretation may occur when tests are performed by inexperienced or inexpert laboratory personnel.(1, 2) False-positive tests are rare, however, and usually can be identified by additional testing.

The HIV-antibody test has limitations. Antibodies usually appear within 3 to 6 months after exposure to and subsequent infection with HIV. Because an infected person does not develop antibodies immediately, a negative result cannot rule out recent HIV infection. If recent exposure is suspected, the test must be repeated in 6 months.

In some settings, HIV RNA testing (bDNA or RT-PCR) may be available and is indicated to diagnose primary HIV infection prior to seroconversion.(17-21) Because of the relatively low specificity of these tests (approximately 95 to 97% for the bDNA), they should not be used as a general screening tool. The false-positive rate will have less impact when used in a relatively high prevalence population, e.g., in persons with signs and symptoms of acute HIV infection who report a recent risk behavior for acquiring HIV. In such circumstances, both an antibody and HIV RNA test should be obtained.

New technology using a less sensitive ELISA, called the detuned assay, is able to diagnose individuals who have already seroconverted on a standard ELISA but are still early in infection. Current HIV antibody tests used in blood banks, the third generation ELISA, detects antibodies approximately 1 to 2 weeks before detectable HIV RNA is present.(22) Commercially available ELISAs detect antibody approximately 2 to 4 weeks after RNA would be detected (M. Busch, personal communication). Thus, for general screening, the benefit of slightly earlier detection is outweighed by the risks associated with a high rate of false positives.

Predictive Value of Tests

The predictive value of HIV antibody tests depends on the prevalence of HIV infection in the population. This concept is crucial to planning testing strategies and interpreting test results. In a population with a very low prevalence of HIV infection, the predictive value of a positive test is low - that is, a positive test result is very likely to be a false-positive. Thus, HIV-antibody testing of low prevalence populations (e.g., all applicants for marriage licenses) is likely to produce more false- than true-positive results.

A positive HIV-antibody test in a person without apparent risks or symptoms, or in a low prevalence population, should be followed by retesting of new serum specimens with appropriate confirmatory tests in a laboratory known for its quality control and proficiency. Laboratories testing for HIV commonly follow one or two positive ELISAs with a confirmatory Western Blot before reporting a positive result of an HIV antibody test.

A rapid single-use HIV-1 ELISA test approved by the FDA is now available. The Single-Use Diagnostic System (SUDS; Murex Corp.) HIV-1 test, which can be used in doctors' offices and clinics, is a screening test only and positive results must be confirmed by standard protocols.(23) The available home HIV tests (Confide; Direct Access Diagnostics and Home Access; Home Access Health Care Corp.) include confirmatory testing of initial positive enzyme immunoassays (EIAs) by either Western Blot or IFA.

Counseling

The Centers for Disease Control (CDC) has published recommendations for HIV testing and counseling.(24) Testing for evidence of HIV infection should always be accompanied by pre- and post-test counseling. The counselor should be knowledgeable about this process. Because results of tests for HIV infection have profound consequences and raise many questions, persons should give written informed consent for the testing procedure and should understand the choices implied by the test results. Counseling should include information about the test, HIV infection, and AIDS, as well as risk behaviors associated with the transmission of HIV. Discussion about the consequences of a positive or negative result for the person being tested (medical care, pregnancy, employment, insurance) and others (family, lovers, friends), as well as the need for appropriate follow-up in the event of positive test results should also be addressed. Explanation of equivocal results that require additional tests may be necessary. Even for a person whose result is negative,counseling is recommended to allay a false sense of security and to promote future risk-reduction behaviors. If the clinical suspicion of infection is high, based on risk behaviors and/or symptoms, follow-up testing should be recommended. In the case of suspected primary infection, referral to a research site or a knowledgeable HIV health care provider should be made even if the antibody test is negative.

Professionals who request HIV antibody tests should be familiar with these counseling issues. Institutions in which the test is commonly administered may consider establishing a testing service staffed by personnel able to effectively counsel, provide follow-up, and appropriately refer individuals seeking testing. In most cases, however, primary care providers will perform this function and can do so effectively. Pre- and post-test counseling and disclosure should be appropriately documented in the medical record according to institutional guidelines.

Both of the FDA-approved home testing systems include referrals to trained counselors for those with positive and indeterminate test results. "Confide" informs those with HIV negative results by recorded message. "Home Access" informs most who are antibody negative by recorded message; some are referred to a counselor. Thus, personal counseling regarding negative results, risk reduction, and follow-up are not generally available with the home testing systems. Discussions about rapid home testing have enhanced the debate about the relative value of counseling and increased access to testing given the realities of frequent poor quality counseling and desirability of home testing.(7)

Confidentiality and Anonymous Testing

Although attempts to maintain confidentiality should be institutionalized and respected, it is misleading to suggest that confidentiality can be guaranteed. In reality, confidentiality may be violated in many ways, including: discussion among personnel acquainted with a clinical case or a laboratory result; access to records by those other than the person ordering the test; discussion about a case when several services are consulted or the case is presented at a conference; inferences drawn by observers such as friends, family members, other patients, or employers; subpoena of records for a legal proceeding; and requirements for reporting AIDS and/or HIV to public health departments.

Anonymous testing systems were developed to guarantee confidentiality. The person being tested is identified in records by a code, such as a number, so that even the person administering the test cannot associate the result with a name. The person tested is then free to divulge or keep secret the test result. Anonymous testing may be reasonable if the test result is not expected to be positive or if fears of violated confidentiality keep a person from obtaining the test. It may also serve as an emotional bridge for people at high risk who are not yet willing to be tested in a confidential setting, but who will benefit from the knowledge of their test result and the associated counseling and referrals.

Anonymous testing is not sufficient for medical management because clinicians must usually have documentation of results in the medical record before instituting diagnostic and treatment interventions. Almost all entitlement programs (e.g., Medicaid, worker's compensation) require official written documentation of test results. Anonymous testing is usually provided in a testing program associated with but separate from the clinical setting.

Home HIV Testing

Several home testing kits are now commercially available. Although some individuals who may not have otherwise been tested may now choose home HIV testing, this option raises concerns about:

In spite of initial concerns and objections to the licensing of home HIV test kits, it is now generally believed that the potential benefits to individuals and society outweigh the potential risks with home HIV testing.

Rapid HIV Testing

Rapid HIV testing will allow clinicians and patients to make decisions about medical management more quickly. It also provides a useful research tool. Not yet fully explored, however, are the counseling issues raised by the possibility of providing an immediate positive screening test result, which necessitates a follow-up confirmatory test. Current counseling strategies have been predicated on the delayed test results. New strategies will need to be developed to counsel those with positive rapid test results. It is crucial to remember that the rapid test is a screening test with a relatively high false-positive rate. Providing results of an unconfirmed screening HIV test to patients is unprecedented in the United States.

Saliva- and Urine-Based Tests

A saliva-based test that collects and stabilizes oral mucosal transudate has FDA approval and is available in the United States for use in doctors' offices and clinics. Its reported sensitivity and specificity are excellent (99.5% and 99.5%, respectively).(25) Both screening EIA and confirmatory Western Blot are performed on the saliva-based specimen. The turn-around time is 3 days.

A urine-based test is also available for office-based screening only. Its sensitivity and specificity are somewhat less than that of the serum- and saliva-based systems (98.7% and 99.1%, respectively).(8) Positive results must be confirmed with a serum-based Western Blot.

Testing to Benefit the Person

The potential benefits of testing include the following:

Potential risks of a negative test result include a false sense of security if the person was recently infected and had not yet produced detectable HIV-antibody as well as the incorrect assumption that a negative test implies immunity from future infection.

Testing May Provide Benefit to the Partner

With the development of more effective treatments for HIV and its associated infections, the debate about names reporting and partner notification has gain come to the forefront in many public health settings. Although many are concerned that names reporting would result in a reluctance for individuals to seek testing, many HIV testing and counseling programs do provide assistance in partner notification.

Testing to Assist in the Medical Management of Asymptomatic Patients

HIV infection includes a spectrum of illness ranging from an asymptomatic state to symptomatic AIDS and severe immunosuppression. Research indicates that eventual progression to AIDS and death is the rule rather than the exception. Fortunately, new therapies may slow this progression and significantly delay HIV-associated morbidity and mortality. To achieve the greatest benefit, however, infected persons must be identified as early as possible. In addition, preventing the spread of HIV infection is best achieved if infected persons recognize their status and its potential risk to others.

Anti-HIV therapy can be beneficial in asymptomatic patients with a detectable viral load and/or a CD4 lymphocyte count below 500 cells/mm³.(10, 11) For patients with CD4 lymphocyte counts below 200 cells/mm³, survival is significantly increased by prophylaxis against Pneumocystis carinii pneumonia. Those with fewer than 50 CD4 cells may benefit from the reduced morbidity and mortality associated with Mycobacterium avium complex (MAC) prophylaxis. To derive these benefits, HIV-infected patients must be identified and their CD4 counts and viral loads determined periodically. Knowledge that HIV infection is present changes the recommended management of several medical conditions, including:

Because of all these medical benefits, widespread testing of asymptomatic persons with a history of high-risk behaviors is appropriate. Some authorities recommend voluntary testing and counseling for all adult patients in hospitals with one or more patients with newly diagnosed HIV disease per 1,000 discharges per year. Offering HIV testing to all women in high seroprevalence areas may be particularly important because many women may not recognize their risk and yet may have acquired HIV infection from a sexual partner.(31)

Testing to Assist Medical Management in Symptomatic Patients

The considerations listed above also apply to symptomatic HIV-infected patients. In many clinical situations, HIV infection is suggested from the history of risk behaviors and clinical manifestations of immunosuppression. Differential diagnosis and management decisions will be influenced by the knowledge of a patient's HIV status. Although initial management decisions can be made on the basis of a history and physical examination consistent with HIV infection, an HIV-antibody test must be used to definitively confirm or rule out infection. The CDC surveillance definition lists a number of conditions considered diagnostic for AIDS only when they are accompanied by laboratory evidence of HIV infection.(32)

Confirmation of a patient's HIV infection influences a number of medical management issues. For example:

Preoperative HIV Testing: No Benefits to Patients

If surgery accelerates HIV disease, then patients might benefit from preoperative screening for evidence of HIV infection, particularly in the case of elective procedures. No data exist, however, to support this hypothesis. Thus, routine preoperative screening is not justified on the grounds of patient benefit.(33) The issue of preoperative screening for the benefit of operating room personnel is discussed subsequently.

Testing to Benefit Public Health and Safety: General Considerations

The public benefits of widespread testing for HIV infection include screening of blood and organ donors to ensure recipient safety and epidemiologic monitoring to apply to the design of prevention and treatment strategies. There is no evidence that widespread testing with the goal of keeping a data base of all HIV-infected persons helps to control the spread of infection, although many public health organizations support names reporting. Prevention and behavior-modification strategies targeting high risk populations do not require learning the identities of individual HIV-infected persons. In the absence of mandatory names reporting, an argument can be made for voluntary partner notifications of those who have tested HIV positive in order to offer screening, treatment, and risk-reduction counseling.

Testing the blood supply has reduced the risk of transfusion-associated HIV infection to a very low level.(4, 5) Rarely, a blood product may be infected and escape detection if the donor was recently infected and has not yet produced detectable HIV antibodies. Some persons might be tempted to donate blood to obtain the results of the blood bank's HIV test. This is clearly undesirable because it encourages donation by potentially infected persons. Free, anonymous, or confidential alternative testing sites discourage this phenomenon.

Testing One Person for the Benefit of Another: General Considerations

Knowing that one person is or is not HIV infected may alter the choices and options of another person. The following scenarios describe some examples:

In these cases, tremendous anxiety could be allayed by knowing that the source's blood had tested negative for HIV antibody. When the source does not or cannot consent to testing, locale-specific laws and policies must be addressed on a case-by-case basis.

Hospitals, police departments, and other institutions employing personnel at risk for occupational exposure should establish easily accessible protocols for testing sources of exposure as well as for testing and following up on those potentially exposed to HIV. Counseling regarding the indications for, and risks and benefits of, post-exposure prophylaxis should also be immediately available.

Testing One Person for the Benefit of Another: Pregnant Women

A large, randomized, well-controlled clinical trial, ACTG 076, has demonstrated clear efficacy in reducing perinatal HIV infection with the use of prenatal and intrapartum zidovudine in HIV-positive pregnant women and their newborns.(34, 35) Voluntary testing of pregnant women with the goal of offering therapy to the woman and her newborn is now widely advocated.

Testing a Person for Institutional Purposes Not Related to the Person's Welfare

In some situations, persons are asked to submit to an HIV test that does not benefit them and may even be detrimental. For example:

Testing of Preoperative Patients to Reduce Risk to Health Care Workers

The following discussion assumes that HIV infection is not used as a reason to withhold medical treatment from a patient. This assumption is fundamental to the ethical, professional, and legal responsibilities of licensed health care workers.

Screening preoperative patients to reduce the risk of infection of health care workers assumes: (1) there is sufficient risk of transmission to justify HIV testing and (2) there are methods to reduce the risk that depend on knowledge of HIV status. Routine screening of preoperative patients for HIV infection cannot be justified by existing evidence that fails to support either of these assumptions. Screening of selected patients may be useful if knowing the patient's HIV-antibody status would lead to a change in medical management.

Screening all preoperative patients has been proposed(36) and discussed.(37, 38) Inevitably and with some regularity, surgical gloves and, less frequently, the surgeon's skin, are punctured during surgery. Certain types of procedures (e.g., orthopedic procedures where fragmented bone must be manipulated in the wound) theoretically have a higher risk for skin puncture, although studies to date have failed to demonstrate an increased risk for HIV infection associated with such procedures. Similar concerns apply to emergency department personnel and others involved in invasive and bloody procedures.

Surgeons have a risk for hepatitis B seropositivity that is 1.5 times higher than other physicians, raising the concern that they may be more likely to become infected with HIV as well.(39) The risk of HIV infection, even after a needle stick exposure from an HIV-infected patient, is estimated to be 0.3 to 0.4%.(6, 40) The risk of hepatitis B seroconversion is 25% after exposure to hepatitis B-infected blood.(37)

The risk of exposure to HIV can be reduced by following recommended universal body substance infection control precautions.(38) These are effective in reducing exposure to HIV, hepatitis B, and other infectious diseases. Knowing that a patient is HIV antibody negative may create a false sense of security because persons may be infected but have not yet produced antibodies.(38) Moreover, other infectious agents capable of causing fatal diseases (such as hepatitis B virus, hepatitis C virus, and HTLV-I) may be present. Thus, infection control procedures should be followed irrespective of a patient's known HIV antibody status.

In some cases, surgeons change management strategies based on knowledge of a patient's HIV infection. If alternative surgical procedures are possible, the one with the least risk to operating room personnel should be chosen (e.g., less hand-to-hand instrument passing, stapling instead of hand-suturing, and electrocautery instead of scalpel cutting).(38) Some of these modifications may, however, prolong operating time, resulting in increased risk to the patient. Testing certain high-risk patients may be useful if a pretest plan exists for using the test result to alter management. The patient should have the ultimate right of refusal of testing without compromising the quality of medical care.

Additional data from prospective studies are needed to evaluate risks of specific procedures and practices. To avoid potential ethical and legal issues, institutional protocols and guidelines should be developed for ordering preoperative testing. We believe that preoperative HIV testing requires pre- and post-test counseling and written informed patient consent.

Testing of Health Care Workers to Reduce Risk to Patients

The evidence and issues of risk to patients from infected health care workers are reviewed in other chapters. Briefly, genetic fingerprinting indicates that the office procedures of a Florida dentist with advanced HIV disease infected five of his patients with HIV.(41-43) No HIV transmission occurred in more than 19,000 other patients who received care from other HIV-infected providers whose practices were reviewed by the CDC.(44) The theoretic risk associated with invasive procedures performed by an infected provider is extremely low; a modeled risk estimation by the CDC suggested that transmission could occur in one of 2.4 million to one of 24 million surgical procedures (H. Kessler, personal communication).

Because of the case of the Florida dentist as well as the case of HIV transmission from an infected orthopedic surgeon in Europe (H. Kessler, personal communication), this issue became widely publicized and led to CDC recommendations that health-care providers who perform "exposure-prone" invasive procedures be tested for HIV and hepatitis B e antigen.(45) This highly controversial position is opposed by a number of professional organizations. Its implementation varies by state law and local hospital policy. The Gay and Lesbian Medical Association's Medical Expertise Retention Program (MERP) was established to assist HIV-infected health care workers in job retention and placement. It is unclear if this issue will be pursued by federal or local health departments.

Testing for Pre-Employment Evaluation

There is no evidence that HIV testing is useful for pre-employment screening. Many HIV-infected persons have no symptoms and no work limitations. Whether a person with HIV infection can work should be determined on the basis of his or her specific signs, symptoms, and functional status rather than by a positive HIV test. The mechanisms of HIV transmission make it clear that HIV-infected persons are not a risk to co-workers or the public unless their job involves a high-risk practice such as sexual intercourse.(46)

Testing for Insurance and Health Care Eligibility

Testing is increasingly required for eligibility for health plans or insurance coverage. The issues involved are complex and include fundamental considerations about how such services are funded and controlled, and the role of the private versus the public sector. The loss of private insurance and health benefits by an HIV-infected person places an increasing financial burden on the public sector to provide for their care.(47)