There are approximately 900,000 people in the United States infected with the human immunodeficiency virus (HIV); these patients receive primary care services from both generalists and subspecialists. Internists also play a vital role in testing, counseling, and educating both HIV-negative and HIV-positive patients. It has become increasingly clear that antiretroviral therapy should only be prescribed by those expert in its use, and thus this chapter will focus on HIV testing, harm-reduction counseling, primary HIV infection and post-exposure prophylaxis. More detailed discussions can be found in the resources, references and websites listed at the end of the chapter.

Testing at-risk people for HIV is essential in order to provide adequate health care and counseling. With no vaccine expected in the near future, behavior modification and treatment of infected patients are the only possible ways to stem the epidemic. Risk-reduction behavior - such as abstinence, condom use and the use of clean needles - clearly reduces HIV transmission. The use of highly active antiretroviral medications makes patients live longer and feel better – it may also make them less infectious. Although testing is increasingly frequent, more than 30 percent of the HIV-infected people in the U.S. are thought to be unaware of their serostatus, and there are an estimated 40,000 new infections a year in this country. While HIV testing can be emotionally challenging for doctor and patient, identifying HIV-infected people has enormous potential benefit both for the individual and for the community.

"High risk" patients are those who belong to populations known to have a high prevalence of HIV infection: sexual partners of patients with HIV, injection drug users, people with multiple sexual partners, men who have sex with men, people with other sexually transmitted diseases, and children of HIV-positive mothers. Blood products have been screened for HIV-1 since 1985 and for HIV-2 since 1992; the risk of infection from blood transfusion subsequent to 1985 is extremely low (estimated at 1/100,000 transfused units). Of note, screening of blood products in the Dominican Republic is not standardized and the risk of a transfusion in the DR is currently unknown.

While complete seroprevalence data are lacking, the rates of HIV in New York City are clearly high. The NYC Department of Health estimates that there are between 100,000 and 140,000 HIV-infected New Yorkers, of whom only 75,000 are aware of their diagnosis; there were 48,145 New Yorkers known to be living with AIDS at the end of 2001. Some surveys suggest that 15 percent of patients in NYC emergency rooms are HIV-infected. We practice in an endemic area.

There is a growing sense that the threshold for HIV testing has been unreasonably high. In addition, generalists fail to recognize common HIV-associated conditions, such as the pathognomonic oral hairy leukoplakia, and may not realize that others – such as seborrheic dermatitis, psoriasis, and herpes zoster are indications for HIV testing. Primary care providers may also fail to identify risk behaviors. Several articles have urged primary care providers to "think HIV" and to abandon what the authors feel is a passive approach to HIV testing. Table 1 summarizes testing recommendations:

TABLE 1: Which adult New Yorkers should be tested for HIV?*

HIV testing involves important public health and privacy issues which neither patient nor provider should take lightly. Counseling (pre-test and post-test) is mandated by law in the state of New York, and the test may not be performed on an adult without written informed consent. Providers should document that they provided pre-test and post-test counseling. Table 2 summarizes recommendations for counseling.

Physicians and patients should be aware of the difference between anonymous and confidential testing. In anonymous testing, samples are given an identifying code and the test center cannot connect the result with a name. The advantage of this method is that it ensures patient privacy. The disadvantage is that it cannot be used to guide therapy; a treating physician must document HIV test results in the medical record. In confidential testing, privacy is a goal that is not always attainable. Test results are available to anyone with access to the medical chart, and HIV is a reportable condition in New York City. Contact tracing and partner notification is also a standard of care – and required by law. Anonymous testing is not available through New York Presbyterian Hospital, but there are several centers in the city which provide this service. Patients can call the NYC AIDS hotline at (800) 825-5448, the New York State AIDS Information Service (800) 541-2437, Planned Parenthood 274-7200, the Geffen Center 367-1100, or GMHC 807-6655 to ask about sites near them. It is important to establish that patients actually received the results of their HIV tests; 10 to 15 percent of patients tested do not return for results.

TABLE 2: Guide to HIV pre-test and post-test counseling*

As noted in Table 2, the essential parts of pre-test counseling are obtaining informed consent to perform the test and preparing the patient to face the results – whether negative or positive. Unfortunately, discrimination against HIV-infected persons still occurs and it can be helpful for patients to think about plans in advance, and to consider carefully whom they will tell if the test returns positive.

Telling patients that they have a potentially fatal disease is one of the hardest things a physician must do. Counseling a patient whose HIV test is positive is a complex task, and one that may take several sessions. While there is no "right way" to do this, the CDC and the NY State Department of Health have published guidelines for HIV counseling. In addition to conveying the basic information about a patient’s HIV status, there are several topics that must be covered in post-test counseling:

1. Behavior modification: Whether the test is negative or positive, patients should be counseled about behaviors that spread HIV. It is important to be specific when discussing sexual practices, injection drug use and behavior modification, even with patients who have been counseled before. Patients should understand that HIV has been isolated from blood, semen, and vaginal secretions, as well as breast milk, pre-ejaculate and saliva. The best way to avoid contracting the disease or giving it to others is to avoid sharing body fluids with others – e.g. to abstain from penetrative sexual intercourse, exchange of secretions or needle-sharing. In lieu of this, harm-reduction strategies (see below) can effectively reduce the risk of transmission.
2. Contact tracing: By law, a patient who is HIV-positive is required to inform sexual and needle-sharing partners of that fact. In practice, leaving notification up to patients is ineffective. Other options include notifying partners yourself, or utilizing the NYC Contact Notification Assistance Project (212) 693-1419, which provides anonymous partner notification. New York State law explicitly permits a physician to violate patient confidentiality and inform patients’ partners of their exposure. Initial steps, however, should include encouraging patients to tell their partners and offering any assistance they require. If patients refuse to cooperate, providers are in a very difficult situation. Informing their partners (if they could be found) could help stop the spread of a deadly disease and allow them to be treated if infected; it will also likely destroy the patient-physician relationship. While partner notification is the responsibility of the primary care provider, as a rule this is not a situation that should be handled by housestaff alone.
3. Pregnancy: It is vital that your patients understand that a pregnant woman infected with HIV can pass the virus on to her fetus. Although antiretroviral therapy in combination with Caesarian section can dramatically reduce the risk of vertical transmission, it does not eliminate it. HIV can be transmitted by breast milk, and HIV-infected mothers should be advised not to breast feed their children.

Recommendations about risk-reduction education and counseling are bounded by two facts: it is clear that behavior modification can reduce the risk of HIV infection, but it is not clear that counseling can dependably cause behavior modification. Studies of specific counseling interventions and meta-analyses of these studies show mixed results. Characteristics of successful counseling programs include culturally sensitive and patient-specific counseling, repeated and consistent counseling over time, and promotion of patient self-efficacy.

As noted in Chapter 19, a careful history is the first step in risk-reduction counseling. Not only is it not helpful to tell a lesbian patient to use condoms or to assume that an elderly patient is abstinent, but such missteps will close further avenues of discussion between patient and provider. A complete sexual history should always be taken, and patient should be asked directly about high-risk behavior. In a respectful and non-judgmental manner, providers should ask about frequency of sexual activity, number and gender of partners, type of penetrative sexual activity and use or non-use of condoms. Patients should also be asked if they are current or former injection drug users.

A "single dose" of counseling is unlikely to effect long-lasting behavior change. Providers should routinely and repeatedly discuss safer sex with sexually active patients. The message should be clear, non-judgmental, patient-specific, explicit and upbeat. The only ways to eliminate the risk of HIV infection are to be abstinent or to have monogamous sex with an HIV-negative partner, and all counseling should include this information. Patients who choose to be sexually active should know that there are specific acts which are highly risky – receptive anal intercourse with an HIV-infected partner being the most dangerous. While exact data are elusive, insertive anal intercourse, "rough sex" (i.e. sex in which mucosa are torn), and sex while genital ulceration is present are also clearly high-risk. Receptive vaginal intercourse appears to be more dangerous than insertive vaginal intercourse; it is also possible (although much less likely) to acquire HIV through oral sex. Prompt treatment of other sexually transmitted diseases is essential, since these are biological "co-factors," which increase the risk of HIV transmission.

Sexually active patients should know that correct and consistent use of barrier protection can dramatically reduce the risk of HIV infection. Latex condoms (male or female), latex dental dams and non-porous (non-microwaveable) plastic wrap are all effective barriers. In order to use a condom correctly, a specific sequence of events must occur: recognition that sexual activity is going to occur, access to a condom, negotiation of condom use with one’s partner, and technical efficacy (removing the condom from its package, putting it on correctly, using it during intercourse and removing it correctly). Only water-based lubricants should be used with latex condoms; oil-based lubricants can weaken the latex. Medicaid will pay for condoms with a prescription.

Users of injection drugs should be referred to detoxification and rehabilitation programs and encouraged to stop. As long as a patient is using, s/he should use clean needles or clean needles and "works" in bleach. Specific guidelines are beyond the scope of this chapter, but questions can be referred to the social workers and peer educators in the Harkness 6 Infectious Disease clinic (305-3174).

Primary HIV infection (P-HIV) also called the "acute retroviral syndrome," is the constellation of symptoms and signs that occurs in most patients as they acquire HIV. It is critically important for generalists in high-prevalence areas to be able to identify this syndrome for three reasons: symptomatic patients will present to their primary care providers, patients with primary HIV are extremely infectious, and early treatment may alter the course of HIV disease.

There are at least 40,000 new cases of HIV in the United States each year, half of which occur in people under 30 years of age. Both prospective and retrospective studies suggest that two-thirds of patients are symptomatic at the time of seroconversion and that most seek medical attention. Symptomatic seroconversion has been documented in all risk categories, but primary HIV infection is rarely suspected, even among high-risk patients. While there is a broad spectrum of severity, the classic presentation is an acute-onset self-limited "flu-like" illness that occurs two to six weeks after exposure and resolves after one to two weeks. Symptoms are felt to correlate to an initial burst of viremia and the associated immune response. Clinical features are listed in Table 3, and are disappointingly nonspecific.

Because the signs and symptoms of the acute retroviral syndrome are nonspecific, patients are often thought to have other viral infections, including mononucleosis, influenza or viral hepatitis. A morbilloform or maculopapular rash on the trunk, arms or face is highly suggestive in the right context. A high index of suspicion and a careful risk history are essential; patients with no exposure to HIV in the past two months do not have primary HIV infection.

Diagnosis of primary HIV is complicated by the fact that patients will not yet have antibodies to HIV; standard ELISA and Western Blot tests may be negative for the first three to six weeks after symptoms occur. If P-HIV is suspected, antigen testing (i.e. RNA viral load testing) should be performed; this should be done with caution as false-positive testing has been reported. We strongly recommend subspecialty consultation with the Infectious Diseases service for patients suspected of having primary HIV infection. If you are considering the diagnosis of primary HIV, the patient should be aware that s/he is likely extremely infectious. Viral load during the acute retroviral syndrome can be "off the chart," and patients should use latex barrier protection or abstain from penetrative sexual intercourse while symptomatic.

Primary care providers in HIV-endemic areas should be familiar with indications for post-exposure prophylaxis (PEP). There are excellent websites dedicated to this topic, and a 24-hour CDC hotline for physicians (888) HIV-4911. One New York City program that provides free PEP has a 24-hour hotline for patients (212) 358-2400. In brief, there are four circumstances in which PEP is usually considered: occupational percutaneous exposure, rape, consensual sexual exposure and exposure through injection drug use/needle sharing. There are limited data with which to make decisions about the last three situations, but PEP guidelines have been developed via expert consensus. It is important to remember that there are not enough data to be dogmatic about PEP, and that in all four scenarios, there are concerns other than HIV infection, including hepatitis and pregnancy.

1. Occupational percutaneous exposure: There is a single case-control study of post-exposure prophylaxis among health care workers exposed via needlestick or laboratory injury. The authors reviewed 33 cases and 665 controls, looking at risk factors for seroconversion and efficacy of zidovudine (AZT) monotherapy. The average risk of seroconversion after injury with an HIV-infected needle is about 1 in 300 (0.3 percent). The study found four independent risk factors for seroconversion: deep injury, injury with a device visibly contaminated with the patient’s blood, procedures involving a needle placed in the source patient’s vein or artery and terminal illness in the source patient. An injury with any one of these characteristics is considered to be high risk. The authors also found that the adjusted odds ratio of seroconversion after PEP was 0.19 – i.e. that the use of zidovudine monotherapy after occupational exposure reduced the risk of HIV infection by 80 percent. Based on these data, the recommendations for PEP in this setting are outlined in Table 4. Of note, the CDC suggest using four weeks of combination antiretroviral therapy (zidovudine + epivir) instead of AZT monotherapy. In specific high risk situations, or ones in which the source patient is already on antiretroviral therapy, protease inhibitors or other nucleoside analogues are recommended. Questions about PEP can be referred to the Infectious Diseases service; speed is essential as there are theoretical reasons to believe that there is a "window of opportunity" for effective PEP. Ideally, persons will take PEP within hours of occupational exposure – the medications are available from Employee Health, the Chief Residents’ office and the Emergency Department.

TABLE 4: Post-exposure prophylaxis for percutaneous injuries

Source status	Class I Asymptomatic; known low titer	Class 2 AIDS; symptomatic infection	Class 3 Pre-terminal AIDS; acute seroconversion; known high titer
Exposure level I – superficial injury	Offer	Recommend	Strongly encourage
II – visibly bloody device; device used in artery or vein	Recommend	Recommend	Strongly encourage
III – deep/ IM injury; actual injection	Strongly encourage	Strongly encourage	Strongly encourage

2. Injection drug use: There are no trials of PEP in the context of needle-sharing, although the pathophysiology of infection is presumably the same as that of percutaneous exposure. It is clear that sharing needles with an HIV-infected person is one of the highest risk behaviors there is. The probability of HIV infection in this setting is estimated to be 0.67 – 1.00, and it is likely that PEP would significantly reduce this risk. However, there are no recommendations to provide PEP in the context of ongoing behavior and the subject is an extremely controversial one. In addition, some experts feel that offering PEP to injection drug users sends a mixed message and may promote IDU itself. There are no Federal guidelines; the UCSF PEP program recommends PEP in this context only if the needle-sharing was an isolated event and the patient intends to abstain from drug use or to use clean needles in the future (see Table 5).
3. Sexual assault: There are scarce data on the risk of HIV-transmission after a single sexual exposure, and few human studies of PEP in this context. In animal studies, PEP can prevent SIV (simian immunodeficiency virus) infection in macaques after mucosal exposure. If the source is HIV-infected, specific sexual acts are riskier than others: receptive anal intercourse > insertive anal intercourse > receptive vaginal intercourse > insertive vaginal intercourse >>> oral sex. Most guidelines recommend offering PEP to survivors of rape.
4. Consensual sexual exposure: As noted, there are no experimental data regarding PEP and sexual exposure to HIV. As with needle sharing, there are those who worry that providing PEP in this context will create a "mixed message," that high-risk sexual behavior is permissible, or that it will even promote risky sexual behaviors. This concern is entirely theoretical, and it should be noted that the availability of the "morning after" pill, an analogous intervention, does not seem to increase risk behavior with regards to pregnancy. In certain communities, such as San Francisco, provision of PEP to patients after (and sometimes even before) high-risk sexual behavior is the standard of care. If the exposure is an isolated event, it may be worth considering PEP. If the risk behavior is ongoing, however, PEP is contraindicated.

TABLE 5: UCSF Guidelines for the use of PEP

Consider PEP if conditions 1-5 are met: High risk exposure (in descending order of risk) unprotected receptive anal intercourse sharing needles or drug paraphernalia unprotected receptive vaginal intercourse unprotected insertive vaginal intercourse unprotected insertive anal intercourse unprotected receptive fellatio with ejaculation Partner is known to be HIV-infected, or in an HIV risk group, or patient was raped Exposure is an isolated event (i.e. broken condom) or patient intends to avoid future exposure through abstinence, safer sex or clean needles Patient presents for care within 72 hours of exposure Patient desires treatment and agrees to adhere to the treatment regimen

While there are many useful textbooks of HIV medicine, the field is a rapidly-changing one and texts may be out of date by the time they are published. The websites listed here have more up-to-date information, but may not be peer-reviewed: caveat emptor.

Websites: http://

1. www.hopkins-aids.edu
2. www.hivatis.org.
3. www.cdcnac.org
4. www.thebody.com
5. www.chp-health.org
6. www.aidsnyc.org/network/access/index.html