HIV infection: Risk factors and prevention strategies

INTRODUCTION — Although the incidence of HIV has dropped considerably since the height of the epidemic, millions of new HIV infections occur globally each year. An understanding of the risk factors for HIV infection lies at the foundation of successful preventive strategies, which should combine both behavioral and biomedical interventions to reduce HIV infection risk.

This topic reviews these risk factors and prevention strategies. More detailed discussion on particular prevention strategies, including pre- and postexposure prophylaxis for HIV, and prevention of mother-to-child transmission of HIV are found elsewhere. (See "Management of nonoccupational exposures to HIV and hepatitis B and C in adults" and "Management of health care personnel exposed to HIV" and "Antiretroviral selection and management in pregnant individuals with HIV in resource-rich settings" and "Prevention of vertical HIV transmission in resource-limited settings" and "HIV pre-exposure prophylaxis".)

MODES OF ACQUISITION — HIV infection is acquired through sexual intercourse, exposure to infected blood, or perinatal transmission. The distribution of the modes of transmission of HIV infection varies in different countries. In the United States, male-to-male sexual contact and injection drug use (IDU) account for more than half of cases [1]. The estimated lifetime risk of an HIV diagnosis in the United States is 1 in 6 for all men who have sex with men (MSM), 1 in 2 for African American MSM, 1 in 4 for Latino MSM, 1 in 23 for women with IDU, and 1 in 36 for men with IDU, in contrast to 1 in 241 and 1 in 473 for heterosexual women and men, respectively [2,3].

In contrast, in some resource-limited areas, penile-vaginal intercourse is responsible for the majority of HIV infections, with a smaller but growing proportion of cases among MSM [4]. Stigma remains a major problem hindering the study of MSM in resource-constrained countries; the incidence and prevalence of HIV in MSM in these settings warrant greater attention. (See "Global epidemiology of HIV infection", section on 'Modes of transmission driving the pandemic'.)

RISK FACTORS FOR INFECTION — Risk of HIV infection varies by type of sexual or parenteral exposure (table 1). However, estimates of risk are mostly based on observational studies and are difficult to quantify since transmission risk also depends on other cofactors that greatly enhance (and occasionally reduce) the probability of infection [5-7]. As an example, risk factors for HIV transmission include high viral load in the source patient, and risk factors for HIV acquisition after exposure include sexually transmitted infections (STIs), and lack of circumcision, as well as certain host and genetic factors [8-10].

Infectiousness of source

Viremia in the source individual and acute HIV — For all modes of transmission, a higher viral load in the source individual living with HIV is associated with a greater risk of transmission. Viral load depends on several factors, most importantly whether the individual is taking antiretroviral therapy (ART) and the stage of HIV disease [5,11,12]. Individuals with stable viral suppression demonstrate no risk for HIV transmission [11,13-18]. Conversely, the risk of HIV transmission appears higher during acute infection, when HIV infection is unrecognized and maximal viremia is present [5]. Phylogenetic and modeling studies (further discussed below) have highlighted that transmission from acutely infected individuals contributes disproportionately to the spread of HIV.

In studies of heterosexual couples, those who transmit HIV have higher levels of HIV RNA than those who do not transmit HIV [5,8,19,20]. As an example, in a study of 415 HIV-serodiscordant couples in Uganda, the baseline serum viral load was higher among transmitting partners than non-transmitting partners (90,000 versus 38,000 copies/mL, respectively) [8]. For each log increase in viral load, there was a 2.5-fold increase in the risk of transmission. In contrast, there were no HIV transmission events from partners living with HIV whose baseline viral load was <1500 copies/mL, suggesting a threshold viral level for transmission [8]. Accordingly, higher viral load in the genital secretions also increases the probability of HIV transmission [19]. Similar findings have been described for perinatal HIV transmission. (See "Antiretroviral selection and management in pregnant individuals with HIV in resource-rich settings", section on 'HIV viremia and risk of infant infection'.)

The viral load can be very high in patients with acute HIV infection or in those with advanced untreated AIDS [21]. Acute infection is also associated with increased viral shedding in genital secretions for several weeks after infection compared with the stable lower levels seen in chronic infection [22]. Accordingly, in several studies, the risk of transmission of HIV appears highest from patients with acute HIV infection and higher from patients with late-stage disease (ie, AIDS) compared with chronic infection [21,23-27]. One modeling study estimated that 38 percent of all HIV transmissions in Malawi were attributable to sexual contact with acutely infected individuals [23]. A phylogenetic analysis of patients with acute and early HIV infection, including mostly men who have sex with men, was similarly suggestive of transmission networks during early infection, with evidence of infection with viruses that could be grouped into closely related clusters [24,25]. Additionally, some animal data also support an increase in viral infectivity during newly-acquired infection [28].

Virus in genital fluids — In general, lower levels of blood plasma HIV RNA are associated with decreased genital HIV RNA levels [19]. Several studies have shown that HIV can still be intermittently detected in seminal fluid of some men living with HIV, despite suppression of HIV viremia with ART [29-33]; this has been associated with concurrent STI, such as urethritis [32]. HIV can also often be detected in the vaginal secretions of women despite the use of ART that suppresses replication in blood [34,35]. However, when plasma RNA is suppressed, detectable viral RNA in genital fluids likely does not represent replication-competent virus capable of transmission. Several studies among women and men have demonstrated that treatment-induced viral suppression in the blood eliminates the risk of transmission. These studies are discussed in detail elsewhere. (See 'Treatment as prevention' below.)

Source with unknown HIV status — Often the HIV status of the sexual or injecting partner is unknown. The probability of HIV acquisition from a source person whose HIV status is unknown is approximated by estimating the likelihood that the source has HIV multiplied by the likelihood of transmission. As an example, in the United States, the prevalence of HIV infection in men aged 18 to 39 in the NHANES study (2003 to 2006) was 0.43 percent [36]. Thus, the estimated risk of transmission following a single random episode of receptive vaginal intercourse in the United States would be approximated by multiplying 0.0043 times 0.001 for an estimate of 0.0000043 (ie, 1 in approximately 233,000).

However, specific information about the source may increase the likelihood that he or she is living with HIV (eg, a history of injection drug use). Furthermore, such calculations are completely different in parts of the world, such as many parts of sub-Saharan Africa, where HIV infection is far more prevalent.

Sexual transmission risk factors — The exact risk of transmission of HIV with sexual exposure is incompletely defined [5,6,37]. However, the data do suggest that risk of HIV acquisition varies by type of exposure (ie, sexual acts) [38-43].

The risk of acquiring HIV after rape by an unknown assailant depends on many factors specific to the assault. This issue is discussed in detail elsewhere. (See "Evaluation and management of adult and adolescent sexual assault victims in the emergency department".)

Sexual behavior — The risk of HIV transmission varies widely depending on the type of sexual exposure. In general, exposures that lead to mucosal disruption and bleeding are associated with higher risk than other exposures. Unprotected anal intercourse conveys the greatest probability of sexual transmission of HIV [37]. Other behavioral factors, such as whether condoms were used, the number of sexual partners, or sex under the influence of recreational drugs, also affect the overall risk of HIV infection. As an example, in a study of 3257 men who have sex with men (MSM) in six US cities, risk factors for HIV acquisition included history of a large number of sexual partners, unprotected receptive anal sex with a partner with an unknown HIV serostatus, and use of nitrate inhalants [44]. In sub-Saharan Africa, having a male sexual partner of older age is a risk of HIV acquisition among young women [45].

The risk of HIV transmission for different types of sexual exposure to HIV in discordant couples has been estimated using different methods (generally through modeling studies) and cohorts (table 1) [46]

●Receptive anal intercourse – One transmission per 72 sex acts

●Insertive anal intercourse – One transmission per 900 sex acts

●Receptive penile-vaginal intercourse – One transmission per 1250 sex acts

●Insertive penile-vaginal intercourse – One transmission per 2500 sex acts

●Receptive or insertive penile-oral sex – Zero to four transmissions per 10,000 sex acts

However, these risk estimates make it difficult to understand the magnitude of the HIV pandemic. Clinicians need to be aware that such estimates of risk are often based on studies of monogamous couples, among whom amplifying factors have been treated and repeated exposure may offer as yet unexplained protection from infection. There are scant empiric data on per contact risk of exposure. Additionally, differences in infectivity are also strongly influenced by the presence of other cofactors, such as concomitant genital ulcerative disease [5]. As an example, some models suggest that rates of HIV transmission may be as high as one transmission event for every three episodes of insertive anal sex between a male source with late-stage disease and a susceptible female with genital ulcerative disease [47]. Thus, using a single value for assessing risk of HIV transmission based on route of sexual exposure fails to reflect the variation associated with other important cofactors.

HIV transmission among MSM is generally higher than among heterosexual partners. Among heterosexual partners, male-to-female transmission may be slightly more common (ie, efficient) than female-to-male transmission, as observed for most STIs. The relative risk of male-to-female compared with female-to-male HIV transmission was well illustrated in one study of 563 HIV serodiscordant couples (including 156 females living with HIV and 400 males living with HIV), in which 19 male partners and 82 female partners acquired HIV during the course of observation [48]. This study suggested that male-to-female transmission was 1.9 times more efficient (95% CI 1.1-33) than female-to-male transmission.

Notably, female-to-female sexual transmission of HIV has only rarely been reported [49,50].

Lack of circumcision — HIV acquisition rates among uncircumcised males are higher than for circumcised males. The biologic basis for this observation may be related to a high density of HIV target cells in male foreskin, including Langerhans cells and macrophages [51]. Randomized controlled trials in Africa have demonstrated that circumcision reduces the risk of female-to-male HIV transmission by 50 to 60 percent [52-55], and this benefit is sustained [55]. However, male circumcision of men living with HIV does not appear to decrease the risk of HIV transmission to the female partner, and efficacy in men who have sex with men has not been demonstrated. (See 'Male circumcision' below.)

Sexually transmitted infections — Concurrent sexually transmitted infections (STIs) have been long known to increase the risk of both acquiring and transmitting HIV infection. The increased risk is widely acknowledged for STIs that cause genital ulcer disease but is also described with other STIs [56]. It has also been associated with the change in vaginal flora that characterizes bacterial vaginosis [57,58] and dysbiosis [59]. In particular, risk of HIV acquisition has been linked to specific inflammatory cytokine profiles evoked by STIs [60] and certain bacterial species that may compromise prevention strategies [61].

Several studies have demonstrated an increased incidence of HIV infection among patients with genital ulcerative diseases, such as herpes simplex virus and syphilis [21,62,63]. As an example, in a study of 174 monogamous Ugandan couples with discordant HIV serostatus, the probability of transmission was approximately four times higher in patients with genital ulceration compared to those without [21]. The effect of these infections on HIV risk is discussed in detail elsewhere. (See "Epidemiology, clinical manifestations, and diagnosis of genital herpes simplex virus infection", section on 'HSV-2 and risk of HIV transmission' and "Syphilis in patients with HIV", section on 'Effect of syphilis on HIV'.)

Increased risk of HIV is not limited to STIs that cause genital ulcer disease. In a prospective study of 242 South African women with high-risk sexual behavior, 28 became infected with HIV over 24 months [64]. Presence of a concurrent STI was associated with an increased risk of infection with HIV (hazard ratio [HR] 3.29, 95% CI 1.5-7.2). The incremental risk for HIV infection with Neisseria gonorrhoeae infection was greater than that with other infections (Chlamydia trachomatis, Mycoplasma genitalium, Trichomonas vaginalis). The majority of women with STIs were asymptomatic.

Among people with HIV infection, STIs may increase HIV shedding in genital secretions in spite of suppression of viremia with ART [65-67]. However, viral copies recovered are not likely to be replication competent.

Genetic background — Similarity of HLA-class-I alleles between HIV discordant couples may affect the risk of transmission by resulting in selected viral strains that are more likely to escape the immune containment of the uninfected partner. In a study of serodiscordant couples, sharing of HLA-B alleles was associated with accelerated transmission of HIV after controlling for other variables (HR 2.23, 95% CI 1.52 to 3.26) [10] (see "Human leukocyte antigens (HLA): A roadmap"). The most profound effects are mediated by the homozygous deletion of the CCR5 receptor [68]. People with this genetic variation are resistant to HIV because the CCR5 receptor is critical to HIV acquisition. Two patients who underwent bone marrow transplants with CCR5 deleted stem cells were cured of HIV infection, demonstrating the critical importance of this receptor [69-71]. (See "The natural history and clinical features of HIV infection in adults and adolescents", section on 'Alterations in the CCR5 coreceptor'.)

Hormonal contraceptive use — Early studies raised concern that certain hormonal contraceptive methods, specifically depot medroxyprogesterone acetate (DPMA), might increase the risk of HIV acquisition. However, subsequent data suggest no increased risk of HIV among women using DPMA or other contraceptives. These studies are discussed in detail elsewhere [72]. (See "HIV and women", section on 'Risk factors for HIV acquisition'.)

Other factors — Studies observing higher incidence of HIV acquisition during late pregnancy and the postpartum period suggest the possibility that physiologic changes during pregnancy affect susceptibility to HIV infection [73,74]. There has also been increased interest in defining specific vaginal fluid cytokine profiles that are associated with and could serve as a marker for increased risk of HIV infection [75,76].

Bloodborne transmission risk factors — The risk of transmission of HIV infection following inadvertent exposure varies widely depending upon the type of exposure.

The risk of HIV infection has been estimated for different types of bloodborne exposure to a source with HIV (table 1) [7,77,78]:

●Blood transfusion – nine infections per 10 exposures

●Needle or syringe sharing – one infection per 150 exposures

●Percutaneous needle-stick – one infection per 435 exposures

●Mucous membrane exposure to blood (eg, splash to eye) – one infection per 1000 exposures

●Other exposure (eg, human bite) – one infection per 25,000 exposures

Studies on needlestick injuries in the health care setting have identified certain characteristics of injury that are associated with increased risk of HIV infection [79]. These include a deep injury, injury with a device that was visibly contaminated with the source individual's blood, injury with a needle that had been placed in a vein or artery, and terminal illness in the source individual. The volume of blood contributes to the probability of transmission since the reported risk of HIV acquisition after transfusion with a contaminated unit of blood ranges from 88 to 100 percent [7]. With the current screening procedures for donated blood, this risk is vanishingly small (see "Blood donor screening: Laboratory testing", section on 'HIV-1 and HIV-2'). In the United States, between 2000 and 2013, there has only been a single confirmed case of occupationally-acquired HIV infection among health care personnel, in a laboratory technician who had a needle puncture while working with live HIV culture [80].

The risk of HIV may be higher among individuals who inject drugs than is reflected by the risk of needle sharing. Many individuals who use injection drugs also participate in sexual behavior, such as unprotected sex and sex with multiple partners, that incrementally increases the risk for HIV infection [81].

Perinatal transmission risk factors — This is discussed in detail elsewhere. (See "Antiretroviral selection and management in pregnant individuals with HIV in resource-rich settings" and "Prevention of vertical HIV transmission in resource-limited settings".)

EFFICACY OF PREVENTION STRATEGIES

Postexposure prophylaxis — For individuals who have a discrete high-risk exposure to HIV, postexposure prophylaxis with an antiretroviral-based regimen is an effective strategy to reduce the risk of infection. This is discussed in detail elsewhere. (See "Management of health care personnel exposed to HIV" and "Management of nonoccupational exposures to HIV and hepatitis B and C in adults".)

Pre-exposure prophylaxis — For individuals who are at ongoing high risk for HIV infection, pre-exposure prophylaxis with an antiretroviral-based regimen is an effective strategy to reduce the risk of infection. Pre-exposure prophylaxis is discussed in detail elsewhere. (See "HIV pre-exposure prophylaxis".)

Treatment as prevention — Treatment as prevention refers to the concept that successfully treating HIV with antiretroviral therapy (ART) minimizes the risk of transmission by decreasing the HIV plasma viral load, and it assumes the potential of a population level benefit.

Several studies have indicated that the risk of transmission to an uninfected sexual partner is negligible when an individual living with HIV has achieved durable viral suppression with ART [82]. This was best demonstrated in a large trial (HPTN 052) that included 1763 HIV serodiscordant heterosexual couples from 13 sites in 9 countries in Africa, Asia, South America, and North America [12,83]. The serodiscordant couples were randomly assigned to an "early ART" arm (initiation of HIV treatment at enrollment) or "delayed ART" arm (initiation of HIV treatment when the CD4 count dropped to less than 250 cells/microL or after an AIDS-related illness). ART consisted mainly of zidovudine, lamivudine, and efavirenz. After 1.7 years of follow-up, interim analysis demonstrated that early ART substantially reduced the risk of HIV transmission to the partner compared with delayed ART (1 versus 27 linked transmissions, as determined by viral sequencing; hazard ratio [HR] 0.04, 95% CI 0.01-0.26). Following these results, ART was offered to the participants living with HIV in the delayed ART group who had not yet initiated it; by the end of the study, 96 percent of the delayed group had started on ART. After a median of 5.5 years and 8904 couple-years of follow-up, the prevention benefits were sustained:

●A total of 46 linked transmissions were identified. Three occurred in couples originally assigned to the early ART group compared with 43 in the delayed ART group (93 percent risk reduction with early ART). A total of eight linked transmissions occurred after the partner living with HIV had initiated ART, but they either occurred within three months of ART initiation or in the setting of ART failure. Thus, no linked transmission events occurred when the partner living with HIV had achieved stable viral suppression on ART.

●There were an additional 16 HIV infections in each group. Of these, 14 in the early ART group and 12 in the delayed ART group were determined by viral sequencing to be genetically unlinked (ie, the source was not the individual's documented partner). An additional six infections (two in the early ART and four in the delayed ART group) could not be sequenced for technical reasons.

●Overall, the incidence of HIV infection in the study was low (0.9 percent per year overall and 2.2 events per 100 person years in the delayed ART group at the interim analysis), which potentially reflects other risk reduction strategies in this study. All participants received condoms and risk reduction counseling. Complete adherence to condom use was independently associated with a reduced risk of both linked and unlinked HIV infections.

The efficacy of ART in reducing the risk of HIV transmission to an uninfected sexual partner has also been supported by many observational studies among heterosexual discordant couples [13,14,84-92]. As an example, in a large retrospective study of almost 39,000 such couples in China, the incidence of HIV infection among uninfected partners was lower if the infected partner was on ART versus untreated (1.3 versus 2.6 infections per 100 person years, respectively, adjusted HR 0.74, 95% CI 0.65–0.84) [86]. Similarly, in a large population-based prospective cohort study of individuals without HIV in rural KwaZulu-Natal, South Africa, where transmission is predominantly through heterosexual sex, HIV infection risk was lower in communities with high ART coverage compared with areas of low ART coverage [93]. Such ecological studies, which examine relationships between exposures and outcomes at the group level, have a number of methodological challenges, including selection bias, confounding, and assumptions about the time lag of effects, which limit the ability to link the exposure to the outcome for the individual [87,94]. Nevertheless, these results indicate the importance of ART as a prevention strategy.

Although most studies of treatment as prevention have included condom use counseling as part of the preventive strategy, studies evaluating serodiscordant couples who specifically do not use condoms also suggest that successful ART prevents transmission [14-16]. As an example, in a prospective observational study of serodiscordant (548 heterosexual and 340 men who have sex with men [MSM]) couples who chose not to use condoms and among whom the partner living with HIV was virally suppressed on ART (<200 copies/mL), there were no documented intra-couple transmission events after more than 1200 couple-years of follow-up [14]. One heterosexual and 10 MSM partners acquired HIV infection during the study period, but viral sequence analysis suggested that these infections were not transmitted from the long-term partner living with HIV.

The efficacy of treatment as prevention is also reflected in the reduction in the rate of mother-to-child transmission of HIV with successful ART of the mother during pregnancy and delivery. (See "Antiretroviral selection and management in pregnant individuals with HIV in resource-rich settings", section on 'Efficacy of ART in preventing transmission' and "Prevention of vertical HIV transmission in resource-limited settings", section on 'Individuals not on ART prior to pregnancy'.)

Data on the efficacy of ART to prevent transmission in populations at risk for other modes of transmission (ie, MSM or injection drug users) are more limited:

●Several observational studies have suggested that ART prevents sexual transmission among MSM [14-16]. As an example, in a study of nearly 800 serodiscordant MSM couples in Europe, of which the partner living with HIV was virologically suppressed (viral load <200 copies/mL) on ART, there were no linked transmissions over 1500 couple-years, which included over 76,000 condomless sex acts [16]. Over a third of men without HIV reported sex outside the partnership, and 15 incident unlinked infections were detected. Similarly, a study of nearly 350 MSM serodiscordant couples in Australia, Thailand, and Brazil did not detect any linked transmission over 232 couple-years, during which the infected partner was on suppressive ART (with a viral load <200 copies/mL) [15].

●In a study of people who inject drugs in Indonesia, Vietnam, and the Ukraine, accelerated ART and substance use management reduced mortality and reduced HIV transmission among injection partners compared with standard of care (0 versus 7 incident infections, which were likely transmission events) [95].

Test and treat — Given the impact of ART in eliminating HIV transmission between serodiscordant couples, there has been substantial interest in a strategy of early ART initiation upon diagnosis as a way to reduce community HIV incidence. Modeling studies have suggested that this "test-and-treat" strategy could significantly reduce HIV transmission in high prevalence countries [96-98], but randomized trials have yielded mixed results, with most demonstrating modest decrease in HIV incidence between intervention and control arms [99-102]. Trial interventions have included combinations of enhanced HIV testing, linkage to care, ART initiation, and adherence support. The magnitude of benefit at the population level requires complex interpretation. Some illustrative trials include the following:

●In a trial in rural Uganda and Kenya (SEARCH Trial) that included over 150,000 participants, communities were randomly assigned to annual HIV testing with rapid ART initiation for those who test positive (intervention) compared with baseline HIV testing and ART administration according to national guidelines (control) [101]. All testing was performed in the setting of multi-disease public health fairs that also addressed hypertension and diabetes mellitus. The intervention resulted in higher rates of viral suppression (79 versus 68 percent at three years) and reduced mortality among individuals living with HIV (relative risk 0.77). However, it did not reduce the three-year cumulative incidence compared with the control group (0.77 versus 0.81 percent, relative risk 0.95, 95% CI 0.77-1.17); across both groups, annual HIV incidence decreased by 30 percent over three years. Marked improvement in the standard of care available to the control group over the course of the trial likely compromised the differences between groups.

●In a trial in Zambia and South Africa (PopART Trial) with greater than 48,000 participants, communities were randomly assigned to standard care or a prevention intervention of annual home-based HIV counseling and testing, linkage to care and adherence support for those who test positive, and encouragement of male circumcision [102]. Those in the prevention intervention group were further randomly assigned to universal ART or ART initiation according to local guidelines. Both intervention arms had increased rates of viral suppression compared to the standard care arm; however, only the intervention arm with ART provided per local guidelines had a lower HIV incidence compared with standard care (adjusted rate ratio for HIV incidence 0.70 [95% CI 0.55-0.88]). The reduction in HIV incidence with the prevention intervention and universal ART compared with standard care was lower in magnitude and not statistically significant (adjusted rate ratio 0.93 [95% CI 0.74-1.18]). When the two intervention arms were combined, there was a 20 percent decrease in HIV incidence relative to control communities. The explanation for the difference in HIV incidence reduction between the two prevention intervention groups is unclear.

●A trial in rural and semi-urban areas in Botswana (Ya Tsie Trial) that included over 12,000 participants also demonstrated a 35 percent reduction in HIV incidence (0.59 versus 0.92 percent per year, incidence rate ratio 0.65, 95% CI 0.46-0.90) with annual HIV testing, enhanced linkage to care with early ART initiation, and encouragement of male circumcision [100].

Further analyses of cost-effectiveness and projected population impact are also ongoing. These studies emphasize the degree of difficulty in treating the right people at the right time with the right drugs to get the maximal benefit of treatment as prevention. Untreated acute and early HIV transmission and migration of people off ART can be expected to lead to HIV transmission at a rate that would compromise the population benefit of a "Test and Treat" strategy. However, population surveys in African countries have demonstrated strong correlation between expansion of treatment of HIV and falling incidence of HIV infection [103].

Condom use — Consistent condom use effectively decreases the risk of sexual HIV transmission and acquisition [104-107]. In 2009, a collaborative statement from the American College of Physicians and the HIV Medicine Association called for wider availability of condoms and education about their proper use to minimize the risk of HIV transmission [107]. In order for condom use to be effective in decreasing HIV prevalence, they need to be used consistently and with ongoing exposures, particularly in areas of high prevalence.

The majority of evidence for condom effectiveness is from observational studies. In a meta-analysis of 12 studies of HIV heterosexual serodiscordant couples, condom usage was classified in three categories: always, sometimes, or none [104]. HIV infection rates were much higher in those who never used condoms versus those who always did (6.7 versus 0.9 infections per 100 person years). A subsequent analysis of these studies suggested that condoms are 90 to 95 percent effective when used consistently [108]. A similar Cochrane review estimated that the consistent use of male latex condoms, defined as using a condom for all acts of penetrative vaginal intercourse, reduces HIV incidence by 80 percent, based on several longitudinal female-to-male and male-to-female serodiscordant couples [106]. Additionally, in a trial of ART to reduce HIV transmission within serodiscordant couples, a self-report of "100 percent use" of condoms compared with "less than 100 percent use" was associated with decreased HIV transmission (HR 0.35; 0.14-0.88) [83]. These clinical studies are consistent with the in vitro finding that latex and polyurethane condoms are impenetrable to HIV viral particles [109].

Since women are often unable to convince their partners to use a condom, there is a need to assess other barrier methods that women can initiate. Female condoms are also impervious to viruses, including HIV; however, there are few clinical data regarding efficacy in prevention of HIV transmission [110].

Male circumcision

HIV infection in heterosexual men — Male circumcision reduces the risk of heterosexual men becoming infected with HIV. Voluntary male medical circumcision is generally safe, but rare cases of tetanus following circumcision have been reported in locations where uptake of infant tetanus immunization is low [111,112]. (See "Tetanus".)

The efficacy of male circumcision to protect that male against HIV infection has been established by several randomized, controlled trials of circumcision conducted in Africa [52-54]. In these trials, men were randomly assigned to an intervention group, in which immediate circumcision was offered, or a control group, in which circumcision was delayed until the end of the study. In all cases, circumcisions were performed by clinicians experienced in the procedure and were rarely associated with moderate or severe adverse effects. These three trials were all ended early because of evidence of reduced HIV incidence in the intervention groups:

●In a trial in Orange Farm, South Africa, circumcision reduced the risk of HIV infection by 60 percent among 3274 men aged 18 to 24 years (20 versus 48 new infections in the circumcision and control groups, respectively, after a mean follow-up of 18 months) [52]. The investigators controlled for behavioral factors, condom use, and health-seeking behaviors.

●In a trial in Kisumu, Kenya, circumcision reduced the risk of HIV infection by 53 percent among 2784 men aged 18 to 24 years (22 versus 47 new infections in the circumcision and control groups, respectively, after a median follow-up of 24 months) [53].

●In a trial in Rakai, Uganda, circumcision reduced the risk of HIV infection by 51 percent among 4996 men aged 15 to 49 years (0.66 versus 1.33 new infections per 100 person years in the circumcision and control groups, respectively, over 24 months) [54]. Following trial closure, the majority of uncircumcised men underwent circumcision. After five years, circumcision remained associated with a lower HIV risk (0.5 versus 1.93 new infections per 100 person years in circumcised and uncircumcised men, respectively) [55].

Population studies in Africa following wider availability and encouragement of voluntary medical male circumcision have also suggested an association between circumcision and reduced incidence of HIV infection and have continued to demonstrate safety of the practice [92,113-115].

Data from the United States are limited. One study examined visit records of heterosexual African-American men who underwent HIV testing while attending STI clinics in Baltimore from 1993 to 2000 and analyzed the association between circumcision and the risk of HIV infection [116]. "Known" HIV risk was defined as patient notification by either their sexual partner or by an intervention specialist from the partner notification system of recent HIV exposure. Among 394 visits by patients with known exposure, circumcision was associated with lower HIV prevalence (10.2 percent versus 22 percent; adjusted prevalence rate ratio, 0.49 [95% CI 0.26-0.93]). In the United States, however, the majority of sexually-acquired HIV infections are among men who have sex with men (MSM), among whom circumcision has uncertain benefit with regard to HIV risk. (See 'HIV infection in MSM' below.)

Any policy to promote circumcision to protect against HIV infection needs to take into account cultural and human rights considerations, the risk of complications from the procedure performed in various settings, the prevalence of infection, and the potential to undermine existing protective behaviors and prevention strategies that reduce the risk of HIV infection [53,117]. Results from observational studies of men who underwent circumcision suggest that this prevention intervention is not necessarily offset by an adverse behavioral impact [118,119]. One study compared sexual behaviors of 324 recently circumcised and 324 uncircumcised Kenyan men at 1, 3, 6, 9, and 12 months after study enrollment [118]. Circumcised men did not engage in more risky behaviors.

HIV infection in women — Although male circumcision reduces HIV acquisition in men, studies have not demonstrated a reduction in HIV acquisition among female sexual partners of men living with HIV who undergo circumcision. Results of these studies suggest that HIV transmission from men to their sexual partners reflects the HIV concentration in genital secretions (ie, semen) rather than exposure to the uncircumcised glans penis.

In a trial from Uganda, 922 men living with HIV with CD4 cell counts ≥350 cells were randomly assigned to immediate versus delayed circumcision for 24 months [120]. Circumcision did not reduce HIV transmission to the female sexual partners without HIV; over a 24-month period, the cumulative probability of female acquisition of HIV was 22 percent in the intervention group and 13 percent in the control group (adjusted HR 1.49, 0.52-3.57). In addition, excess HIV transmission occurred within the first six months in the male circumcision arm, particularly among those who resumed intercourse prior to wound healing. These findings suggest sexual abstinence or condom use should be strongly advised until surgical recovery (estimated at six weeks).

HIV infection in MSM — To date, no randomized controlled studies of male circumcision for HIV prevention have been conducted among men who have sex with men (MSM). Observational studies have suggested a possible but uncertain protective effect. Because some (but not all) MSM engage in both insertive and receptive anal intercourse, it is difficult to conduct studies to show a benefit of circumcision in this setting.

In a meta-analysis performed of 15 observational studies among 53,567 MSM (52 percent of whom were circumcised), the odds of having HIV were 14 percent lower among the circumcised men, but the difference was not statistically significant [121].

Counseling and harm reduction strategies — Risk-reduction counseling and harm reduction strategies can reduce behavior that results in higher risk of HIV infection. Risk-reduction counseling ranges from high-intensity behavioral discussion tailored to an individual's risk to brief prevention messages to group-based strategies [122,123].

Individuals report greater condom use and fewer sexual partners with behavioral risk-reduction interventions, and some studies report that counseling decreases risk of sexually transmitted infection (STI), including HIV [124,125]. The precise efficacy of risk-reduction counseling may depend, in part, on how it is conducted. (See "Prevention of sexually transmitted infections", section on 'Prevention counseling'.)

For injection drug users, harm reduction interventions, such as voluntary opioid substitution therapy and needle exchange programs, can reduce risky injection behavior. Opioid substitution therapy is associated with decreased illicit opioid use, injecting use, and sharing injection equipment [126]. Needle exchange or supervised injection programs have also been associated with decreased needle reuse and sharing, safe syringe disposal, and more hygienic injection practices [127]. Although these strategies have not been evaluated in high-quality trials, observational and modeling studies suggest that they are associated with decreases in HIV infection [126,128,129]. Treatment of opioid addiction with buprenorphine/naltrexone is also associated with increased use of and adherence to ART among individuals living with HIV [130]. As noted above, enhanced ART and substance abuse management have demonstrable personal and public health benefits [95].

Treatment of sexually transmitted infections — "Classical" sexually transmitted infections (STIs) that cause ulcers or mucosal inflammation are strongly associated with HIV transmission [56]. However, it has been difficult to reliably demonstrate the benefits of STI treatment for HIV prevention [131]. Screening and treatment of STIs in at-risk individuals are important for general sexual health. (See "Screening for sexually transmitted infections", section on 'Introduction'.)

Population-based studies have generally been unable to show a decrease in HIV incidence with treatment or prevention of STIs [132,133]. It is generally concluded that STI treatment to prevent HIV infection requires treatment of just the right STI pathogen, at just the right time, with effective agents, goals that are not readily achieved. Additionally, the findings in these studies may reflect problems with methodology, patient adherence, and overlapping interventions in the control group in these studies as opposed to a true lack of preventive benefit with STI treatment [131].

Experimental approaches — A vaccine to prevent HIV infection has been an elusive goal of HIV research. Despite several candidates, an effective HIV vaccine has yet to be developed [134-136]. However, new approaches continue to be investigated [137,138]. In addition, discovery and characterization of antibodies that neutralize a large number of strains of HIV have generated hope that such antibodies can be delivered or generated in some way to provide long-acting prevention from HIV [139-143].

CLINICAL APPROACH TO HIV PREVENTION — A comprehensive approach to HIV prevention includes risk reduction among individuals living with HIV to reduce transmission and among at-risk individuals to reduce acquisition. Biomedical interventions (such as the antiretroviral-based strategies of treatment as prevention and pre- and postexposure prophylaxis, as well as voluntary medical circumcision for uninfected heterosexual males) are proven and highly effective in reducing HIV infection. Nevertheless, behavioral interventions (such as condom use and risk-reduction counseling) remain crucial elements of prevention, and most studies of biomedical interventions included continued emphasis on behavioral modification. While no single behavioral intervention is perfectly effective, in aggregate, it is believed that such interventions have reduced the spread of HIV in some populations. The magnitude of the epidemic without such interventions cannot be predicted.

Accordingly, the following section discusses how certain aspects of biomedical and behavioral interventions are combined for a patient-specific approach. This discussion is generally consistent with recommendations on HIV prevention from multiple expert organizations [144-146]. Discussion of the studies supporting the individual prevention strategies is found elsewhere. (See 'Efficacy of prevention strategies' above.)

Individuals living with HIV — Preventing transmission of HIV from individuals living with HIV relies on engaging and maintaining patients living with HIV in care, initiating antiretroviral therapy (ART) early, maintaining successful therapy, and reducing behavioral risk factors.

Antiretroviral treatment — Suppression of plasma viremia with ART effectively minimizes the risk of sexual [12,14,83,84] and perinatal HIV transmission and likely does the same for other modes of transmission. Thus, ART is a key strategy to curb the spread of HIV. (See 'Treatment as prevention' above.)

In the United States, ART initiation is recommended for all patients living with HIV, regardless of CD4 cell count, to decrease AIDS and non-AIDS associated morbidities and mortalities [144,146]. Prevention of HIV transmission is an additional rationale for the more widespread use of ART. For resource-limited settings, the World Health Organization (WHO) has also recommended ART initiation for all patients living with HIV, regardless of CD4 cell count or clinical stage, given the mounting evidence of the clinical benefit of ART, even at high CD4 cell counts [147,148], and the reduction in the risk of transmission to uninfected partners with successful ART [149]. (See "When to initiate antiretroviral therapy in persons with HIV" and "Use and impact of antiretroviral therapy for HIV infection in resource-limited settings", section on 'Recommendations from the World Health Organization'.)

Successful administration of ART relies upon public health measures to optimize timely linkage of newly diagnosed individuals living with HIV with medical care and retaining them in care. Treatment success also relies on patient adherence to medication, which can be facilitated by reductions in pill burden, reduced dosing frequency, and ongoing counseling to assess and encourage ART adherence.

Counseling on condom use — All patients living with HIV who are not on stable ART should use condoms to reduce the risk of sexual transmission. For individuals living with HIV with durable viral suppression on ART, the risk of transmission to an uninfected sexual partner is negligible and use of condoms for the purpose of preventing HIV transmission is no longer considered essential [12,84,150]. The extensive data supporting treatment for prevention has led to the "U=U" campaign, inspiring individuals living with HIV to recognize that "Undetectable" HIV in the blood as a result of successful ART leads to an "Untransmissible" status [17]. (See 'Treatment as prevention' above.)

Condoms do have other benefits, however, and we continue to advise their use in certain situations, including:

●For patients at risk for other sexually transmitted infections (STIs), which are common among individuals living with HIV regardless of ART, and which condoms effectively prevent [151]. (See "Prevention of sexually transmitted infections", section on 'Male condom use'.)

●When a partner living with HIV has loss of virologic suppression, in which case condoms can prevent against HIV transmission.

Risk reduction — Routine assessment of ongoing risk behavior should be performed, and counseling on reducing those risks should be individualized.

For injection drug users, addiction treatment (eg, with opioid substitution or with buprenorphine-naltrexone) and participation in needle exchange programs, if available, are additional useful strategies to decrease risk behavior and, likely, HIV transmission. (See 'Counseling and harm reduction strategies' above.)

Finally, we screen and treat STIs in individuals living with HIV given the increased risk of STIs in this population, the association of STIs with HIV transmission, and the benefit of treating STIs beyond potential HIV prevention [151]. (See 'Treatment of sexually transmitted infections' above and "Screening for sexually transmitted infections", section on 'Patients with HIV infection'.)

Individuals at risk for HIV

Identifying new infections — Identifying individuals living with HIV is crucial to directing interventions to prevent transmission. Thus, HIV testing is a major element of comprehensive preventive strategies. In particular, acutely infected individuals, who often have exceptionally high levels of viremia, have been identified as major sources of transmission (see 'Viremia in the source individual and acute HIV' above). Testing strategies should employ diagnostic tests that are highly sensitive for acute infection and should include routine screening for all individuals with repeat testing of individuals with high ongoing risk for HIV. (See "Screening and diagnostic testing for HIV infection".)

Once HIV infection has been diagnosed or excluded, preventive efforts for that patient can be individualized. (See 'Individuals living with HIV' above and 'Risk reduction' below.)

Risk reduction — In uninfected individuals, preventing acquisition of HIV focuses on reducing behaviors that increase the risk of infection and utilizing antiretroviral-based prophylactic strategies. In areas where the prevalence of HIV is high and sexual transmission is primarily among heterosexuals, voluntary medical circumcision is another important strategy.

All uninfected patients should be assessed for their risk of HIV infection, and counseling on reducing those risks should be individualized.

●For all patients at risk for HIV infection, we continue to advise consistent condom use. (See 'Condom use' above.)

We also recommend screening and treatment of STIs in individuals at risk for HIV given the shared risk factors for HIV and other STIs, the association of other STIs with HIV infection, and the benefit of treating STIs beyond potential HIV prevention [151]. (See 'Treatment of sexually transmitted infections' above and "Screening for sexually transmitted infections", section on 'Screening recommendations'.)

●For injection drug users, addiction treatment (eg, with opioid substitution or buprenorphine-naltrexone) and participation in needle exchange programs, if available, are additional useful strategies to decrease risk behavior and, likely, HIV acquisition. (See 'Condom use' above and 'Counseling and harm reduction strategies' above.)

●For those who have high ongoing risk for HIV infection, pre-exposure prophylaxis effectively reduces the risk of infection. (See "HIV pre-exposure prophylaxis".)

●For those who have had a mucosal or parenteral exposure to HIV within the prior 72 hours, postexposure prophylaxis with an antiretroviral regimen is associated with a reduced risk of infection. (See "Management of nonoccupational exposures to HIV and hepatitis B and C in adults", section on 'Exposure to HIV' and "Management of health care personnel exposed to HIV", section on 'Post-exposure prophylaxis'.)

Circumcision has demonstrated efficacy in reducing the risk of HIV infection among heterosexual men. WHO and the Joint United Nations Programme on HIV/AIDS (UNAIDS) recommend scaling up voluntary male circumcision as a HIV prevention intervention in several African countries with high rates of HIV and low baseline rates of male circumcision. In the United States and Europe, where sexual transmission among men who have sex with men (MSM) is dominant, circumcision has not demonstrated substantial benefit. (See 'Male circumcision' above.)

Serodiscordant couples — The term "serodiscordant couples" refers to ongoing sexual partnerships between an individual living with HIV and an uninfected individual. Prevention strategies for individuals living with HIV and at-risk individuals are thus relevant to each of the members of the serodiscordant couple, respectively. (See 'Individuals living with HIV' above and 'Individuals at risk for HIV' above.)

HIV transmission can be reduced with ART for the partner living with HIV [12,83] and with pre-exposure prophylaxis for the uninfected partner [152]. For all patients living with HIV in a serodiscordant partnership, we recommend initiation of ART in order to prevent transmission to the uninfected partner (see 'Treatment as prevention' above). No studies have directly compared the two strategies for HIV prevention, but ART for the partner living with HIV has the additional benefit of reducing the AIDS and non-AIDS morbidity and mortality for that individual [83]. Pre-exposure prophylaxis for the partner without HIV may be indicated until the partner living with HIV has achieved stable viral suppression on ART (typically by six months after initiating therapy [153,154]), if ART fails to suppress HIV in the partner living with HIV for any reason, and for risk-taking behavior outside the partnership. (See "HIV pre-exposure prophylaxis", section on 'Sexual risk'.)

Counseling on condom use and other risk reduction strategies is discussed elsewhere. (See 'Counseling on condom use' above and 'Risk reduction' above.)

Options for serodiscordant couples desiring pregnancy are also discussed elsewhere. (See "Use of assisted reproduction in HIV- and hepatitis-infected couples", section on 'Natural conception in serodiscordant couples'.)

Pregnant and breastfeeding women — Prevention of mother-to-child transmission of HIV involves ART for the pregnant woman and postexposure prophylaxis for the infant. Precise management differs in resource-rich and limited settings and is discussed elsewhere. (See "Antiretroviral selection and management in pregnant individuals with HIV in resource-rich settings" and "Prevention of vertical HIV transmission in resource-limited settings" and "Prevention of HIV transmission during breastfeeding in resource-limited settings".)

A PUBLIC HEALTH PLAN FOR THE UNITED STATES — Utilizing the prevention interventions described below, the United States has described a plan to reduce incident HIV in the United States by 90 percent by 2030 [155]. The plan has "4 pillars":

●Diagnose all individuals with HIV as early as possible after infection

●Treat HIV rapidly and successfully to achieve sustained viral suppression

●Prevent at-risk people from acquiring HIV through interventions including the use of pre-exposure prophylaxis

●Rapidly detect and respond to emerging clusters of HIV infection to further reduce new transmissions

This plan has a strong focus in the southern United States and some additional urban centers where HIV spread continues [155].

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: HIV prevention".)

SUMMARY AND RECOMMENDATIONS

●Modes of acquisition – HIV infection is acquired through sexual intercourse, exposure to infected blood, or perinatal transmission. For all modes of transmission, a higher viral load in the source individual living with HIV is associated with a greater risk of transmission. (See 'Modes of acquisition' above and 'Viremia in the source individual and acute HIV' above.)

●Risk of sexual transmission – The risk of sexual transmission of HIV varies widely depending on the type of exposure. In general, exposures that lead to mucosal disruption and bleeding are associated with higher risk than other exposures. Unprotected anal sex conveys the greatest probability of HIV transmission and oral sex the lowest. Lack of circumcision is associated with HIV acquisition in men, and concurrent sexually transmitted infections are associated with both transmission and acquisition. Other behavioral factors, such as whether condoms were used, the number of sexual partners, and sex under the influence of recreational drugs, also affect overall sexual HIV transmission risk. (See 'Sexual transmission risk factors' above.)

●Risk of parenteral transmission – The risk of parenteral transmission of HIV depends on the volume of contaminated blood that an individual is exposed to and the depth of the injury or exposure (eg, mucous membrane exposure, needle-stick, or injection into vessel). (See 'Bloodborne transmission risk factors' above.)

●Prevention strategies – A comprehensive approach to HIV prevention includes behavioral strategies (eg, harm reduction strategies such as condom use, opioid substitution therapy, needle exchange programs), antiretroviral-based strategies, screening and treatment of sexually transmitted infections (STIs), and in some settings, circumcision.

•Behavioral interventions – Behavioral interventions (such as condom use and risk-reduction counseling for drug use) remain crucial elements of HIV prevention. In addition, continued condom use remains useful to reduce risk of other STIs. (See 'Condom use' above and 'Counseling and harm reduction strategies' above.)

•Antiretroviral therapy – Antiretroviral-based strategies to reduce HIV transmission are warranted for those with and without HIV infection.

-Persons with HIV – For individuals living with HIV, successful antiretroviral therapy (ART) is a key strategy to prevent transmission. The risk of HIV transmission from an individual living with HIV who has achieved stable viral suppression to an uninfected partner is negligible. (See 'Individuals living with HIV' above and 'Treatment as prevention' above.)

-Persons at risk for HIV – Antiretroviral-based strategies for persons without HIV can also reduce the risk of HIV acquisition. These include pre-exposure prophylaxis for individuals with ongoing risk, and post-exposure prophylaxis for those who have had a mucosal or parenteral exposure to HIV within the prior 72 hours. (See 'Individuals at risk for HIV' above and "Screening and diagnostic testing for HIV infection" and "Management of nonoccupational exposures to HIV and hepatitis B and C in adults" and "HIV pre-exposure prophylaxis".)

-Approach for serodiscordant couples – For serodiscordant couples, we recommend antiretroviral treatment of the partner living with HIV as the primary prevention strategy instead of antiretroviral prophylaxis for the uninfected partner (Grade 1A). Pre-exposure prophylaxis for the partner without HIV is additionally indicated until the infected partner has achieved stable viral suppression on ART (typically by six months after initiating therapy), if ART fails to suppress HIV in the partner living with HIV for any reason, or for risk-taking behavior outside the partnership. Management of serodiscordant couples desiring pregnancy is discussed elsewhere. (See 'Serodiscordant couples' above and "Use of assisted reproduction in HIV- and hepatitis-infected couples", section on 'Natural conception in serodiscordant couples'.)

•Screening for HIV and other sexually transmitted infections – For at-risk individuals, routine HIV screening and having a low threshold to test for HIV is crucial to identify new HIV infections and engage patients in interventions to prevent HIV transmission.

We also screen for and treat other sexually transmitted infections in individuals with or at risk for HIV infection given the shared risk factors for HIV and other STIs, the association of other STIs with HIV infection, and the benefit of treating STIs beyond potential HIV prevention. (See 'Treatment of sexually transmitted infections' above.)

•Circumcision – Circumcision has demonstrated efficacy in reducing the risk of HIV acquisition among heterosexual men, but the benefit among men who have sex with men (MSM) is uncertain. In areas where the prevalence of HIV is high and sexual transmission is mainly among heterosexuals, such as in several African countries, voluntary medical circumcision can be an effective element of HIV prevention intervention. (See 'Male circumcision' above.)

ACKNOWLEDGMENT — UpToDate gratefully acknowledges John G Bartlett, MD (deceased), who contributed as Section Editor on earlier versions of this topic and was a founding Editor-in-Chief for UpToDate in Infectious Diseases.