HIV diagnosis and testing: what every healthcare professional can do (and why they should)

Authors


Correspondence: Frank Maldarelli, HIV Drug Resistance Program, NCI, NIH, 9000 Rockville Pike, Building 10 Rm. 5A06, Bethesda MD 20892, USA. Tel: 301-435-8019, Fax: 301-480-1735, E-mail: fmalli@mail.nih.gov

Abstract

Over the last thirty years, the human immunodeficiency virus (HIV) epidemic has matured. In the United States, HIV has changed from an explosive outbreak to an endemic disease; currently, an estimated 1.1 million people are infected with HIV, including a substantial number who are unaware of their status. With recent findings demonstrating the high transmissibility of HIV early in infection, and the potential benefit of early initiation of treatment, it is essential to identify as many infected individuals as possible. The Centers for Disease Control and Prevention (CDC) has expanded HIV testing to include any healthcare setting, including dental offices. Testing advances, including oral testing, have reduced the window period of HIV infection. Dental care represents a key, reliable, independent, and confidential link between the healthcare system and the general population that has been under-utilized in the effort to control the HIV epidemic. HIV testing is straightforward, and knowledge of the types of testing will afford dentists an important opportunity to help advance and preserve the health of their patients and to promote the public health of their community. Here, we review the basics of HIV testing and discuss new changes in the approach to HIV diagnostics.

Introduction. HIV: A maturing epidemic

From the time of its initial identification over thirty years ago, the human immunodeficiency virus (Pant Pai et al, 2012) epidemic has continued to expand in the United States and throughout the world. The Centers for Disease Control and Prevention (CDC) estimates that approximately 1.1 million Americans are infected with HIV (prevalence of 0.35%), and of those, more than one in five are unaware of their status (CDC, 2008). Of those newly diagnosed with HIV infection, nearly 40% have advanced disease and progress to AIDS within 1 year, evidence that they may have been infected (and capable of transmitting HIV) for many years prior to their diagnosis (CDC, 2008). Approximately 56 000 new cases of HIV occur yearly (Hall et al, 2008). HIV in these acute infections is more transmissible than chronic infection, probably because of increased HIV levels in blood and secretions, delayed immune responses, and the presence of genital lesions from co-infection with other sexually transmitted diseases (Fideli et al, 2001; Pilcher et al, 2001, 2004, 2007; Miller et al, 2010; Morrison et al, 2010).

Early on in the epidemic, most cases were young, concentrated in large cities, and associated with specific risk factors, especially intravenous drug use and men who have sex with men (MSM); mother-to-child transmission expanded cases to include newborns of infected mothers. After the HIV epidemic expanded in major metropolitan centers, it maintained endemic transmission, likely through a number of factors that are distinct from the initial epidemic ones (Rothenberg, 2007). Over time, the epidemic spread outside of metropolitan areas and expanded to include older age groups (Hall et al, 2005). Currently, overall prevalence of HIV infection varies over 100-fold in the United States, ranging from 0.6 (Vermont) (CDC, 2012) to 130 cases per 100 000 (Washington, D.C.). Rates vary regionally, with the largest rates of urban infections in the northeast and proportionally higher rates of infection in the rural south (Hall et al, 2008) with infections occurring the deep south in minority groups and women (Pence et al, 2007). Many newly infected patients are older, with 10% of new diagnoses in individuals over the age of 50 years (Hall et al, 2008). Women now account for over 20% (2.5/100000) of cases (CDC, 2012).

These sobering epidemiologic characteristics have persisted despite important advances in HIV medicine. HIV testing is simple, rapid, and convenient; screening modalities include oral fluid and urine in addition to blood testing, all of which have high sensitivity. In those diagnosed with HIV, referrals are readily available and therapeutic options are straightforward, flexible, with once-a-day regimens. Medical and social support systems for patients are robust.

A substantial reservoir of undiagnosed individuals, prolonged infection prior to diagnosis, changing demographics, improved diagnostics, and a change in perceptions require new view of the HIV epidemic in the United States and with it a fresh perspective on public health strategies for prevention and control.

In the United States, public health efforts to advance HIV testing have undergone extensive revision over the last 25 years. The demographic and geographic expansion of the US HIV epidemic present new challenges in identifying infected individuals. Early in the epidemic, after the development and licensing of the first HIV test in 1985, testing was limited to blood donation centers, and shortly thereafter, to so-called alternative sites, generally hospitals. These initial approaches were not successful in identifying all patients with HIV infection, and testing policies have undergone sequential expansion to reach larger proportions of the population for testing. During the last 20 years, newer and more flexible testing modalities using oral fluid or urine have been developed, making testing easy, reliable, and rapid, with results available more and more quickly. With the recent approval by the FDA of a home test for HIV, screening has never been more accessible.

Key epidemiologic and scientific advances resulted in successive revisions in recommendations for testing. Improvements in testing modalities contributed to slowing the epidemic, yet the estimated incidence rate of HIV has remained stable. In addition, a substantial proportion of patients identified with HIV infection have advanced disease; these so-called late presenters have likely been infected for many years prior to detection. To address the continued transmission of HIV, CDC has expanded the recommendations for testing to include individuals between the ages of 18 and 64 presenting for routine or emergent health care (CDC, 2012). As such, the newest recommendations expand testing to the broadest possible population. Guidelines have not yet been incorporated nationwide, and routine testing remains the exception not the rule in emergency room settings (Johnson et al, 2011; Hoover et al, 2012) not all healthcare professionals are well versed in the new recommendations (Mohajer et al, 2012) or do not recommend testing according to the guidelines (Berkenblit et al, 2012). Widespread implementation will be essential for overall success of this approach.

In this review, we will summarize the essentials of HIV diagnosis, current testing modalities, with specific focus on issues of interest to the oral surgery and dental community.

HIV Natural History

HIV variants are highly diverse, but all HIV variants are detectable with routine assays. HIV belongs to the lentivirus genus of the Orthoretrovirinae subfamily of the Retroviridae family of viruses (King et al, 2012). As a group, retroviruses are all enveloped (+) strand RNA viruses that replicate through an essential reverse transcription event that converts virion RNA to a double-stranded DNA copy that is integrated into host genomes. Retroviruses are distributed worldwide and throughout a variety of human and non-human hosts.

Human immunodeficiency viruses are divided into two types, HIV-1 and HIV-2, which differ in origin, distribution, pathogenesis, and treatment. HIV-1 comprises the majority of infections worldwide and can be further divided into four groups M (‘Main’, worldwide distribution), O (‘Other’, largely geographically limited to Central and West Africa), N (non-M, Non-O, small number of infections in West Africa), and P (several reported cases in individuals from West Africa) (Buonaguro et al, 2007; Luft et al, 2011; Lihana et al, 2012). The main group M is the principal source of HIV-1 in the world. Infection has been further subdivided into 10 subtypes (A–J); the majority of HIV infections in the United States are of subtype B, although the numbers of non-B subtypes have been increasing as HIV-infected individuals immigrate to the United States. HIV-2 is largely restricted to West Africa, and, to a lesser degree, countries with traditional, economic, or cultural ties with West Africa. There are 8 subtypes (A–H) of HIV-2, although the majority of infections are either subtype A or B. HIV-2 is a lethal infection, but pathogenesis is often slower than HIV-1.

The broad genetic diversity of HIV groups, types, and subtypes has placed particular demands on developing assays with comprehensive performance characteristics capable of detecting all HIV subtypes. Early versions of HIV diagnostic assays did not uniformly detect all subtypes and groups. Over the last 15–20 years, optimization of screening assays has been highly successful; currently, all HIV types, groups, and subtypes are identified with high specificity and sensitivity with HIV-1/2 ELISA assays; rare cases of group N or group P infections were still detected using standard HIV-1/2 ELISA assays (Yamaguchi et al, 2006; Plantier et al, 2009; Vallari et al, 2010, 2011).

Upon acute infection, HIV replication proceeds with the expansion of a single or limited number of HIV variants (Keele et al, 2008; Kearney et al, 2009; Keele and Derdeyn, 2009). Infection may proceed with a non-specific flu-like illness or may be asymptomatic. In the typical course, a short aviremic window period, typically days, is followed by a rapid increase in viral RNA levels in plasma, which peaks at levels as high as 105–106 viral RNA copies/ml plasma or more within several weeks of transmission (Figure 1). Viral RNA levels decline thereafter to a set point, which may remain relatively stable for years (Figure 1). Host immune responses lag behind viral replication but can be detectable as a cellular immune response with HIV-specific CD8 cell responses, or a serologic response with IgM and, subsequently, IgG responses within days to weeks of infection (Figure 1). Antibodies are also detectable in urine and in crevicular fluid, providing convenient modalities for detection and diagnostic screening. Except in unusual cases, serologic responses are life long. As a result, diagnosis of HIV infection using serologic means remains highly reliable.

Figure 1.

Time course of HIV infection. HIV replication is followed by serologic immune responses. The time during which HIV is present but not detectable represents a vulnerable window period for laboratory diagnosis (NAT = nucleic acid testing)

Following acute infection, individuals undergo slow progressive decline in immune function. CD4 cell counts represent a useful marker of immune deficiency: over the course of infection, CD4 cells gradually decline as a function of the viral RNA level, with higher viral RNA levels associated with faster CD4 decline. Immune deficiency is associated with risk of contracting opportunistic infections or developing AIDS-associated neoplasms, which represent frequent causes of death in HIV infection. Many of these severe complications do not occur, however, until immune deficiency is relatively advanced. As a result, individuals may be infected for years with no overt specific signs or symptoms of HIV infection; such individuals represent a continual risk of transmission. Identifying individuals early in infection permits timely engagement of the healthcare system, early administration of antiviral therapy, with benefits of preservation of immune function and prevention of transmission.

HIV Testing

A sequential testing strategy maximizes sensitivity and specificity

The requirements for HIV diagnostics are rigorous. HIV detection assays must be comprehensive to identify genetically diverse HIV variants, sufficiently sensitive to detect low levels of HIV present early in infection, and highly specific to limit false-positive assays (Table 1). Assays must be broadly applicable with robust performance characteristics in settings such as blood donation, where the prevalence of HIV is likely to be low (perhaps 0.01–0.1%), but at the same time, testing must have robust specificity so as not to overburden testing centers with resource-consuming false-positive results, necessitating costly follow-up testing. Thus, highly sensitive and specific assays are necessary for wide dissemination and for diverse applications. Early on, a number of different serologic and virologic assays were evaluated for large-scale applications, but ultimately no single test had sufficiently high sensitivity (few false negatives) combined with high specificity (few false positives). To maximize sensitivity and achieve the highest positive predictive value (true positives/test positives), a two-step detection system was developed, consisting of a screening test with high sensitivity followed by a confirmatory test designed for maximum specificity. Screening, therefore, is designed to identify as many individuals as possible with HIV infection, and with current tests, sensitivity of HIV screening is greater than 99.9%. However, to achieve this level of sensitivity, a significant false-positive screening rate is necessary; thus, all individuals who screen positive are tested by a confirmatory assay, which has a specificity of greater than 99.9%, and such performance characteristics efficiently distinguish true-positive from false-positive screening reactivity.

Table 1. Common HIV testing modalities
TestVirus DetectedAnalyte detectedFDA Approval?Useful for Acute HIV?
HIV-1/2 ELISAHIV-1 and HIV-2Antibody to HIVYesYes
Rapid HIVHIV-1 and HIV-2Antibody to virusYesreduced sensitivity
HIV-1 Western BlotHIV-1 onlyAntibody to HIVYeslimited
HIV-2 Western blotHIV-2 onlyAntibody to HIV-2NoLimited
4th Generation ELISA HIV-1 and HIV-2 Antibody to HIV AND p24 antigen Yes Preferred
HIV RNAHIV-1 onlyHIV RNAYesPreferred
Home Test for HIVHIV-1 and HIV-2antibody to virusYesLimited

Three new assay developments have advanced testing of HIV infection: The development of RNA detection assays and combination p24 antigen and ELISA assays improved detection of early HIV infection, and sensitive multispot assays permitted clarification in identification of HIV-1/HIV-2 dual infection.

No test or testing algorithm is perfect, and it is essential to be aware of limitations in various testing modalities. In general, the chief limitation of any HIV testing is the inability to detect HIV early after HIV infection has occurred. This early period, denoted the ‘window period’ (Figure 1), when the viral load is often high, but the immune system has not yet developed a detectable response, has gradually shrunk over the last 20 years, and current state-of-the-art assays are, in general (but not universally), sensitive enough to detect HIV within 2 weeks of infection.

HIV Screening Assays

Serologic and virologic assays have been developed for HIV screening. The first test to detect HIV, an enzyme-linked immunoassay (ELISA), was introduced in 1985, initially designed to screen blood donations. These early ‘first-generation’ ELISA assays represented technologic breakthroughs that provided a strong measure of public health protection, but had significant false-positive rates, as the source of the antigen used for testing was not extensively purified and the design of the assay, an indirect ELISA, exhibited significant background for a low signal-to-noise ratio. Subsequently, two improvements increased HIV sensitivity and specificity. First, incorporation of HIV antigens expressed from cloned HIV genes for use in ELISA formats reduced non-specific reactivity. Second, the development of so-called ‘sandwich’ ELISA assays, which engage both antigen binding domains of the immunoglobulin molecule, resulted in marked improvements, with decreased background reactivity and increased specificity. These basic improvements, resulting in sequential second- and third-generation assays, are largely responsible for improved sensitivity and specificity, as well as reducing the window period of HIV from several months to several weeks.

The discovery of HIV-2 prompted changes in HIV screening, adding an HIV-2 ELISA and, subsequently, replacing these two assays with a combination HIV-1/2 ELISA. As genetic diversity of HIV was defined and numerous subtypes spread worldwide, ELISA screening assays underwent important developments to improve detection of all subtypes, and currently available assays have undergone rigorous testing to reliably detect all known HIV group M subtypes, as well as HIV group O virus, and newly identified N and P viruses. (Branson, 2007).

To enable more widespread screening and to reduce loss to follow-up, ELISA tests that were rapid or used either urine or saliva were developed. Rapid tests now yield results in less than 30 min, enabling more effective integration of outreach, counseling, education, and testing efforts (Delaney et al, 2011). Rapid tests use response to HIV-1 antigens (usually envelope proteins gp41, gp120, or gp160) from oral secretions or whole blood, both specimens that do not require sophisticated processing, to screen for infection with particle agglutination, immunochromatography, or immunoconcentration techniques.

One commonly used test, OraQuick (available from OraSure Technologies, Inc., Bethlehem, PA, USA), while initially approved for use with finger-stick blood (FDA, 2012), can also be applied to oral secretions as well. Testing takes advantage of the relatively substantial levels of anti-HIV antibodies present in crevicular fluid (Lamey et al, 1996; Granade et al, 1998). The test is easy to use and interpret and has been waived under the Clinical Laboratory Improvement Amendments (CLIA), no longer requiring certification or formal training, enabling its use in smaller clinical and point-of-care (POC) settings rather than large traditional laboratories (Branson, 2007). Sensitivity was reported to be slightly lower when the test used oral fluids rather than finger-stick blood (Pavie et al, 2010), but field testing noted equivalent sensitivity of blood and oral fluid testing (Zachary et al, 2012). Oral testing has been shown to have preserved sensitivity within 5 min of using mouthwash and is likely to retain sensitivity after a dental cleaning. In limited studies, consumption of alcohol, brushing of teeth, use of mouthwash, and smoking tobacco more than 5 min prior to testing were shown to have no effect on test sensitivity (OraQuick package insert).

It has been reported that rapid testing has a lower sensitivity for detecting early infection with HIV, a time when viral loads can be high and risk for transmission is elevated (Louie et al, 2008; Patel et al, 2010; Piwowar-Manning et al, 2011). False negatives have also been reported in patients with advanced disease or on suppressive antiretroviral therapy (O'Connell et al, 2003, 2006); meta-analyses have suggested that a small decrease in sensitivity of rapid oral testing may reduce positive predictive value in low-prevalence populations (Pant Pai et al, 2012).

Screening with p24 assays

Early in the development of HIV, diagnostic kit assays to detect HIV antigen p24 were found to be sensitive and specific for HIV infection, although the greatest sensitivity was early in HIV infection, when plasma viremia was particularly high and the presence of highly avid cross-reacting antibody was relatively low. Approved since 1989, p24 assays have been useful in identifying early HIV infection and, in recent years, have been adapted in combination with antibody ELISA approaches to generate improved ‘fourth-generation’ detection, especially early in infection (Chappel et al, 2009; Daskalakis, 2011).

All reactive screening assays are repeated, and only samples that are repeatedly reactive will also be processed for confirmatory testing. Note that the results of screening are reported as ‘reactive’ or ‘non-reactive’, not ‘positive’ and ‘negative’. As screening only identifies individuals who may be infected with HIV, definitive results are only obtained using confirmatory assays.

Confirmatory testing

In the United States, several confirmatory assays are available, including Western blot testing and immunofluorescence assay, with Western blot testing being the most commonly used (Dewar et al, 2009). Western blotting strategy identifies antibody reactivity to individual HIV virion proteins. Lysates of HIV virions are subjected to sodium dodecyl sulfate–polyacrylamide gel electrophoresis, which uses an electric field to separate proteins in gels as they migrate according to their molecular size. A number of HIV proteins are resolved by this method, including membrane precursor glycoproteins gp160 and products gp120 and gp41, structural protein precursor, p55, viral products p 24 and p17, and enzymes p66, p51, and p31. Proteins in the gel are electrotransferred to nitrocellulose paper, which is then used to react with serum dilutions. Antibodies reacting to individual HIV proteins are identified using anti-human antibodies chemically coupled to enzymes. As a consequence, Western blotting identifies reactivity to individual HIV proteins.

Several algorithms to determine whether reactivity is sufficient to report HIV positivity have been developed. The American Red Cross method requires at least 2 reactive bands for a positive diagnosis. A negative result is reported when there are no positive bands. Results that have only one reactive band are reported as ‘indeterminate’. Indeterminate results may reflect early HIV infection, during which time there is sufficient antibody to generate a reactive ELISA, but not sufficient quantity, type, or avidity to generate a positive Western blot. It is also possible that indeterminate Western blot may simply represent background antibody reactivity in an individual who is truly HIV negative. Early studies demonstrated that indeterminate results in HIV-negative individuals persist, and they are likely to remain Western blot indeterminate over time (Davey et al, 1992).

In the past, resolving indeterminate Western blots was not straightforward; repeat testing over 1–2 months typically identified individuals with early HIV infection, as their Western blot results inevitably became positive with time. Indeterminate results have been reported in patients with autoimmune diseases, pregnancy, or who have received certain vaccines (Dewar et al, 2009), and if subsequent testing for HIV is negative, further investigation may be warranted. Introduction of nucleic acid testing for diagnosis has offered a critical new modality to resolve indeterminate Western blots.

Importantly, FDA-approved confirmatory (Western blot or immunofluorescence) assays are available only for HIV-1. Although there are HIV-2 slot blot and Western blot tests available in the United States, none have undergone the rigorous FDA approval process. Thus, patients with HIV-1/2 ELISA reactive, but Western blot negative results, may be infected with HIV-2 and should be evaluated as such, with judicious and critical use of HIV-2 Western blot detection. Healthcare professionals involved in HIV testing always perform detailed history to evaluate patients to determine potential sources of infection. Diagnostic algorithms for screening and confirmation, for example, Figure 2, (Dewar et al, 2009) can be useful to help guide diagnosis in a stepwise fashion.

Figure 2.

Typical algorithm for diagnosing HIV infection. Diagnosis begins with history and physical examination and proceeds through screening and confirmatory assays (NAT = nucleic acid testing)

Early introduction of antiretroviral therapy can affect HIV diagnostics. In adults, HIV antibody levels may decline after introduction of therapy; although ELISA assay results are generally positive, introduction of therapy prior to development of fully positive Western blot may delay development of positivity or arrest the evolution of the Western blot at the indeterminate stage. In the absence of a positive Western blot, true HIV infection can be ensured by documenting a positive HIV RNA level (Aptima, below) prior to introduction of antiretroviral therapy.

New Developments in Laboratory Diagnosis: closing, but not shutting the window period

Fourth-generation ELISA assays

Fourth-generation HIV antigen–antibody (Ag/Ab) combination assays simultaneously detect HIV p24 antigen and antibodies to HIV-1 (groups M and O) and/or HIV-2. These assays are more sensitive than third-generation assays due to their ability to identify established infections as well as acute infections. During the acute phase of HIV infection, HIV RNA and p24 antigen are detectable, but HIV antibodies are not. HIV p24 antigen detection with fourth-generation assays reduces the window period, allowing for earlier diagnosis of HIV infection.

Two fourth-generation HIV Ag/Ab combination assays are FDA-approved: ARCHITECT HIV Ag/Ab Combo (Abbott Laboratories, Abbott Park, IL) is a chemiluminescent microparticle immunoassay (CMIA) and was approved in June 2010. More recently, the GS HIV Ag/Ab Combo EIA (Bio-Rad Laboratories, Redmond, WA), an enzyme immunoassay (EIA), was approved in July 2011. Both are qualitative assays to detect HIV-1/HIV-2 infection, including acute or primary HIV-1 infection. They do not, however, distinguish between the presence of HIV-1 p24 antigen, HIV-1 antibody, or HIV-2 antibody. Repeatedly reactive specimens must be followed up with an assay that differentiates HIV-1 and HIV-2 antibodies. If neither HIV-1 nor HIV-2 antibodies are present, then the specimen must be tested for HIV-1 RNA because the patient could be in the window period of HIV infection. Importantly, both these fourth-generation assays were approved to aid in the diagnosis of HIV-1/HIV-2 infection in pregnant women and in children as young as 2 years of age.

Nucleic Acid Testing (NAT)

The development of sensitive assays to detect specific nucleic acids has been applied to HIV with great success. HIV RNA quantitation is used to monitor patients with known infection and in response to antiretroviral therapy. These monitoring assays, which have been available since the development of combination antiretroviral therapy, have not been approved for diagnosis because the assays have not been validated for use in low-prevalence populations, and there is a low, but significant, false-positive rate in HIV-negative individuals. As a consequence, new assays were developed to specifically identify HIV in low-prevalence populations. The results of these developments is an RNA amplification assay, Aptima , which identifies HIV RNA using transcription-mediated amplification (Ren et al, 2008; Nugent et al, 2009). RNA is extracted and captured by hybridization to oligonucleotides complementary to highly conserved HIV domains; RNA is collected using magnetic beads, and exogenous reverse transcriptase is used to synthesize a short promoter region, the RNA sequence with short DNA product, which is then utilized as a prime template by T7 polymerase to generate many copies of HIV that are detected in a hybridization protection assay.

The assay has a low limit of detection, c. 10–20 copies, and is ideal to investigate early HIV infection and to resolve indeterminate Western blots. NAT represents the most sensitive indicator of HIV infection and is consistently more sensitive than third- or fourth-generation ELISA assays. As such, Aptima has been approved for screening or for confirmation, but not both for an individual; it must be used in combination with either an EIA or Western blot. In circumstances where ELISA results are reactive, Aptima can be used to confirm the infection. In the case where Aptima results are the first results suggestive of HIV infection, ELISA or Western blot can be performed for confirmation. In using NAT for diagnosis, it is important to use only those assays that are FDA-approved for their specific indication. Using assays other than those approved for diagnosis may yield confusing results.

Home Collection and Home Testing for HIV

HIV testing has moved progressively away from traditional blood donation and healthcare settings and is now available for home use. Home collection services were approved by FDA in 1996, a system in which individuals purchase a collection kit and obtain a confidential and unique identifier by phone, along with pretest counseling. The individual then uses a lancet to obtain a small amount of blood and blots the blood on to filter paper, which is then returned to the company. After the sample is processed, the individual dials a phone number to receive results and post-test counseling. This strategy ensures that, at least to some degree, patients undergo testing with counseling, consent, and confidentiality.

The ease and reliability of rapid and simple tests for HIV infection has stimulated additional interest in home testing. In May 2012, the FDA approved the first true home testing kit, Oraquick In-Home HIV test (FDA, 2012), manufactured by Orasure technologies. This oral test uses a swab to sample crevicular fluid in an ELISA screening assay. False-negative results, especially in acute infection, have been reported, and performance characteristics in otherwise untrained individuals has not been established. As with all ELISA screening assays, the test will require confirmation for all reactive results. It remains uncertain how this approach will ensure consent, but pre- and post- test counseling will be available (Koval, 2012; Mandell, 2012; Paltiel and Walensky, 2012).

HIV Testing Guidelines: Recommendations Expand Faster than Acceptance

As described above, testing recommendations have undergone significant expansion over the course of the last 25 years, driven largely by improvements in technology permitting wider dissemination, the compelling nature of the epidemic, and greater acceptance of infected individuals. Early discussions yielded early recommendations for prevention of infection in specific circumstances, including perinatal and occupational transmission (1986). Early recommendations issued shortly after ELISA testing was available were designed to assist in interrupting transmission and afford opportunities for counseling at-risk populations, defined as MSM/bisexual men, intravenous drug users (IVDU), patients with signs or symptoms of AIDS or AIDS-related complex, individuals from countries with high AIDS rates, commercial sex workers, individuals with hemophilia who received clotting factor products, newborns of high-risk mothers, and sexual partners of at-risk individuals (MMWR, 1986). In the absence of effective therapy, early recommendations emphasized education and counseling to prevent further transmission. By 1987, the CDC took a broader approach and characterized their recommendations by risk groups: those with sexually transmitted diseases, engaging in intravenous drug use, commercial sex workers, self-perception of risk, women at risk through sexual transmission, individuals presenting for medical evaluation (including all hemophiliacs), admitted to hospitals, or placed in correctional institutions, and those planning marriage (MMWR, 1987). The strength of the recommendations to test in individual groups was tempered by the relative prevalence of HIV. In 1993, testing was recommended for inpatients and outpatients at acute care hospitals, and voluntary counseling and testing (VCT) of all individuals between the ages of 15 and 54 was recommended for individuals in regions of the country with higher (≥1%) HIV prevalence (MMWR, 1993). These new guidelines also provided recommendations regarding medical and preventative services.

With time, recommendations became more specific and extensive. By 2001, recommendations were based on broad subject review, and wide distribution, including Internet access for maximum dissemination (MMWR, 2001). Although confidential testing and referral were ‘routinely recommended’, economic and other limitations in universal implementation were clear; as a result, testing was still targeted to regions of the country with higher (≥1%) HIV prevalence, but testing in other regions was encouraged.

Despite expansion and revision of recommendations, testing remained inadequate to control the epidemic, and HIV incidence rates appear to have remained stable over the last 15–20 years. Testing rates remained low even in emergency rooms (3.2 tests/1000 visits) and was driven by clinical presentation (Hsieh et al, 2008); actual counseling often did not conform to guidelines (Castrucci et al, 2002). One stark example is the substantial and relatively constant proportion of newly diagnosed individuals who receive a diagnosis of AIDS within 1 year of diagnosis of HIV infection; approximately 35% of all newly diagnosed patients fall into this category of so-called ‘late testors’. Such individuals have likely had infection for a prolonged period prior to diagnosis; vital opportunities for therapy are lost while the risk of ongoing transmission continues. And, alarmingly, the frequency of such individuals in the United States did not change appreciably, despite broad changes in recommendations for testing. As a consequence, new recommendations, released by the CDC in 2006, recommended additional expansion in HIV testing for all individuals between the ages of 18 and 64, regardless of perceived risk or local prevalence of HIV(Branson et al, 2006).

From the earliest discussions of testing, the effective and ethical implementation of testing featured three requirements: confidentiality, counseling, and consent, essential components of the VCT model. Mandatory testing for HIV has been rejected in the United States and worldwide by the World Health Organization as coercive. There are, however, circumstances when HIV testing has been mandated: court-ordered HIV testing in criminal cases has been upheld in the US justice system, and inmates in federal as well as some state prisons undergo mandatory testing. Most states recommend HIV testing of pregnant women using an opt-out strategy; a complete listing of HIV testing laws is readily available (http://www.nccc.ucsf.edu/).

The concept of consent and strategy of administering VCT has shifted over the last 10 years. Previously, consent had been explicit, often written, but with recent efforts to streamline testing (in high-volume sites such as emergency departments), so-called ‘opt-out’ strategies are employed, in which patients are informed that testing will be performed as routine care unless the patient specifically refuses testing. Many states use opt-out testing approach to HIV testing during pregnancy, which has been largely accepted, but the effects of widespread implementation of opt-out testing have yet to be clearly measured. The CDC conducted a rigorous review and solicited broad input from experts within academia and industry to streamline HIV testing. The most recent recommendations, promulgated in the 2006 CDC Guidelines (Branson et al, 2006), recommend HIV testing as part of routine medical care for everyone between 13 and 64 years of age without regard for relative risk of HIV infection unless the risk is <0.1%(Branson et al, 2006). Screening remains voluntary but employs an opt-out approach to permit broad inclusion strategies. Screening still requires consent, although verbal consent is sufficient. A substantial proportion of patients visiting emergency rooms (ERs) have expressed support for testing (Haukoos et al, 2008), although issues surrounding the additional support necessary to maintain efficient workflow in ER settings have not been resolved. Additional suggestions to support at least partial adoption of the recommendations have been reported (Lyons et al, 2007).

Knowledge and acceptance of new guidelines in medical settings has been variable (Jain et al, 2009; Mahajan et al, 2009), although increased dissemination of guidelines and education programs represent a useful step, as a substantial number of healthcare workers remain unaware of appropriate guidelines (Jain et al, 2009). Broad implementation remains debated, and Holtgrave has reported that a targeted counseling and testing approach is superior to opt-out testing and is more cost-effective (Holtgrave, 2007).

HIV Testing: An Approach for the twenty-first century

Testing for HIV infection is a public health imperative. New recommendations support widespread screening for HIV as part of a routine health care in all settings, including home testing. Healthcare settings are most advantageous for HIV diagnosis; they provide confidence, education, and overall patient support. Healthcare professionals in the dental community occupy a key position in HIV diagnosis because they provide routine care to individuals, many of whom do not otherwise interact with the healthcare system on a regular basis. Pollack et al (2010) reported that 75% of individuals at risk for HIV infection had visited a dentist within 2 years. Feasibility of testing in a dental setting has been clearly demonstrated (Hutchinson et al, 2012). Limitations of screening in dental clinics, including reimbursement, are practical issues that are manageable in the face of the potential benefit, and well-informed and trained health professionals will succeed in improving the health of their patients and communities with comprehensive HIV education, counseling, and testing. Routine testing has broad personal and societal benefits and has the potential to yield substantial reductions in HIV infections.

Author contributions

EW, TT, and FM drafted the manuscript.

Ancillary