The prevalence of anal intercourse among heterosexual adults has not been well-studied in many populations. This knowledge gap is important because, relative to vaginal intercourse, anal intercourse may be as much as 20 times more efficient for transmitting HIV. Given the transmission efficiency of anal intercourse, even a low prevalence of anal sex among heterosexuals could account for a large number of HIV infections .
Estimates from the U.S. and U.K. suggest an average heterosexual anal intercourse prevalence of 11% (range = 2% to 39%) [1, 2]. The highest rates have been found among female sex workers and patients at sexually transmitted infections (STI) clinics . As many as 20% of STI clinic patients in South Africa report engaging in unprotected anal intercourse with their most recent partner . It is also noteworthy that, among heterosexuals, anal intercourse often co-occurs with other risky sexual practices (e.g., multiple and concurrent sex partners, selling/trading sex, using alcohol and other drugs), increasing risk further [3, 5].
The role of anal intercourse in southern Africa - where the greatest concentration of heterosexually-transmitted HIV cases are found - remains under-investigated. Studies of specific population sub-groups are most common. For example, research has shown that 43% of female sex workers  and 20% of truck drivers  report lifetime anal intercourse. In addition, 14% of men and 10% women recruited from South African STI clinics and impoverished communities reported engaging in anal intercourse in the past three months . Among individuals reporting anal intercourse, 28% engaged in this behavior more often than vaginal intercourse. As occurs elsewhere, anal intercourse in South Africa tends to occur in the context of other high-risk factors, including substance use. However, focusing on sex workers and persons who have contracted an STI may overestimate the prevalence of anal intercourse.
Prevalence estimates of anal intercourse can help interventionists effectively tailor prevention messages. HIV prevention programmers must decide the degree to which messaging should emphasize anal intercourse relative to other risk behaviors; the former may be less prevalent but more efficient in transmitting HIV. Ignoring anal intercourse in prevention messages may suggest it is unimportant. On the other hand, overemphasizing a low-base rate behavior in prevention messages may inadvertently reinforce the idea that vaginal intercourse is low risk for HIV transmission. Determining the appropriate balance of attention to relative risks is best guided by behavioral surveillance research of targeted risk practices.
An important group for targeted HIV prevention in South Africa is persons who patronize informal alcohol serving establishments (shebeens). Shebeens are common throughout southern Africa and, importantly, are a meeting place for sex partners [9–11]. Among men and women recruited from informal local shebeens, more than one in four report having met sex partners at a shebeen . Individuals who meet sex partners in shebeens have greater numbers of recent sex partners and higher rates of unprotected intercourse compared to persons who did not meet partners in shebeens . The high frequency and heavy drinking in HIV prevalent places where sex partners are met creates significant risks for the spread of HIV. Thus, informal drinking venues and their surrounding communities are a high-priority for HIV prevention interventions in South Africa [9, 14–16].
Given that shebeens provide an environment where sexual partnerships are formed among patrons who are likely to be infected with HIV, [17, 18] this study focused on the prevalence of anal intercourse in this context. We hypothesized that anal intercourse would be observed at relatively low frequencies among a minority of persons, and that individuals who engaged in anal intercourse would report other high risk practices; a pattern of multiple risk behaviors would suggest that such persons may be at highest risk for HIV infection. We also used mathematical modeling techniques to estimate the increased risk for HIV acquisition conferred by unprotected anal intercourse to women who drink at shebeens.