More than 340 million cases of curable sexually transmitted infections (STIs) including Treponema pallidum, Neisseria gonorrhoeae, Chlamydia trachomatis and Trichomonas vaginalis were estimated to have occurred worldwide in 1995 [1]. Previous studies have shown that the presence of other concomitant STIs increases the likelihood of HIV transmission [2–4].

In Pakistan, the estimated number of people living with HIV (PLHIV) was 96,000 in 2008, with an adult HIV prevalence estimated at 0.1% [5]. The HIV epidemic in Pakistan has thus far followed the so-called Asian epidemic model, where HIV initially affects injecting drug users (IDUs) before passing on to the general population through the bridging populations of men who have sex with men (MSM) and/or women selling sex [6]. There is also evidence of non-concomitant sexual networking among IDUs, women selling sex and MSMs. The 2^nd Generation HIV Surveillance, conducted in 2006-07, revealed that 9.9% of women selling sex reported having sex with IDUs, and 22% of IDUs reported having bought sex from women selling sex [7].

The first national STI study in Pakistan was conducted in 2004 in order to determine the disease burden of HIV and STIs. The study showed a high burden of STIs among women selling sex, with the following prevalences; T. pallidum 7%, N. gonorrhoeae 12%, C. trachomatis 11% and T. vaginalis 19%. The study also highlighted a low level of knowledge about AIDS among the women selling sex; 24% had never heard of AIDS and 26% had not used a condom with their most recent sexual partner (http://www.nacp.gov.pk/library/reports/). In Pakistan, the terms HIV and AIDS are commonly used synonymously. The Ministry of Health, Government of Pakistan, through the National and Provincial AIDS Control Program introduced preventive interventions among the most-at-risk populations in various urban settings in 2005 through public private partnerships. These interventions were focused on provision of curative and preventive services, including syndromic management of STIs, provision of condoms, and behaviour change communication. The interventions were also tied with the 2^nd generation HIV surveillance among most-at-risk population, which estimated the HIV prevalence and its related risk behaviours, but did not measure the prevalence of STIs (http://nacp.gov.pk). Our study assessed both the HIV and STI prevalence with related risk behaviours.

The history and context of sex work among women selling sex in Lahore, Pakistan has been described earlier [8, 9]. Briefly, the Mughal Dynasty, which ruled most of India, Pakistan, Bangladesh, and Afghanistan for over four centuries patronized artists, singers, and dancers by establishing settlements near the palaces where they lived. One such area was established next to the Royal Fort of Lahore. This area in Lahore is still called Heera Mandi (the market of diamonds) or Bazaar-e-Husn (the market of beauty), and is the oldest establishment in Pakistan where women sell sex [9]. Music, singing, dancing, and selling sex was occurring side by side in these areas. Traditionally, the women passed on their profession, including selling sex to their daughters, but not to their daughters-in-law. The latest estimated figure of women selling sex in Lahore is 26,000, which is the highest number of mapped women selling sex in any city in Pakistan (World Bank, 2006). Currently, the women selling sex in Lahore are found to work in at least four different settings or contexts: Brothel based, Kothikhana based, Street based and Home based. In the Brothel based typology, Brothels are fixed venues, which are usually operated by "madams" sometimes together with other network operators. Women selling sex live in the house, which is licensed for singing and dancing. These houses are normally located in larger sex work or red light areas and are stable locations, known by clients. Sex work takes place either at the brothel or at the client's residence and the women selling sex have usually no other work. Kothikhana is an idiomatic expression for a sex work venue that literally means "grand house". However, Kothikhanas are generally small premises, which are rented by a "madam". They are often located in residential areas and their location is usually kept a secret. A key feature is that they move from time to time when the "madam" determines that the location is unsafe or unsuitable. Women selling sex at Kothikhanas work on weekdays from 9 AM to 5 PM and are hired on fixed rate per client by the "madam". In the Street based typology, women selling sex solicit clients in public places such as busy streets, intersections, bus and train stations and marketplaces. Sexual transactions then occur at a venue chosen by the woman selling sex or the client. In the Home-based typology, women selling sex usually live with their families. Contact with clients is established by means of mobile phones or through network operators. Sex work takes place either in the client's home or hotels, or a place provided by the network operator. Sex workers usually work part time, operating when required for financial purposes [7].

The aim of our study was to estimate the prevalence of HIV and STIs and assess knowledge and risk behaviours related to HIV/STI among women selling sex in Lahore, Pakistan.

Among the 730 women selling sex in Lahore who participated in our study, 91% were married, 0.7% were infected with HIV and an additional 18.5% were suffering from different curable STIs. Only 19% and 83% of the participants were aware of the terms HIV and AIDS, respectively, and only 37% had correct knowledge about transmission and prevention of AIDS.

According to the 2^nd generation HIV surveillance report from 12 major cities in Pakistan, the likelihood of the HIV epidemic affecting women selling sex is very high, as, for example 10% to 32% of the women selling sex have sex with IDUs, who have an overall HIV prevalence of 15.8% (CI 95%: 14.7-16.9; range 1-53%) (http://nacp.gov.pk). In another study conducted in two major cities, 24% and 21% of IDUs reported that they had sex with women selling sex [15, 16]. However, in studies conducted among women selling sex, the prevalence of HIV infection has been reported to be less than one percent (http://nacp.gov.pk) [15, 16]. There might still be a window of opportunity to keep HIV prevalence below 5% among women selling sex, if the scaling up of services, including laboratory diagnostics, syndromic management and periodic presumptive treatment of STIs, and increase in condom use, is achieved. Such interventions in India among women selling sex have demonstrated that the package of services cited above has the potential to prevent 22%-35% of all new HIV infections among women selling sex [17].

An earlier study from 2004 examining women selling sex (n = 400) in Lahore recruited through RDS in only one of the four areas, found a substantially higher prevalence of most STIs, i.e. N gonorrhoeae (12.3%), C trachomatis (11%), T vaginalis (19.3%), and T pallidum (7.1%) (http://nacp.gov.pk). However, our study, which was conducted in three different areas of Lahore, and recruited almost the double number of participants, found lower prevalences of STIs as compared to the 2004 study which was done only in one area of Lahore. Our study as well as another study [15] showed the prevalences of different STIs to be less than 10% among women selling sex. A review of 42 studies on STI rates around the world have emphasised that women selling sex tend to have comparatively higher rates of STI than the general population [18]. In Kolkata, India, women selling sex had a high prevalence of STIs, with 34% testing positive for N gonorrhoeae and 23% for T vaginalis [19]. In a study performed in Vietnam the prevalence of N gonorrhoeae and C trachomatis were 14.9% and 48.4%, respectively [20].

The comparatively low levels of HIV and STIs in our study are best understood in the context of what is known about the local sexual and social networks and possibly the transmission patterns. Mathematical models of networks of women selling sex suggest that the prevalence of HIV is extremely sensitive to an increase in condom use and number of irregular partners [21, 22]. In "Area B", where women selling sex were Kothikhana based, we found the lowest reported rate of "Always using condoms", the highest number of clients per woman, and hence the highest prevalence of HIV. Furthermore, the low prevalences of HIV and STIs in our study and that of Hawkes et al in two other cities of Pakistan[15], as compared with other studies, could perhaps be explained by relatively high condom use, relatively low numbers of sexual partners and availability of antibiotic treatment through syndromic management of STIs.

In our study, the reported rate of "always use condoms" was high (65%) as compared with previous studies that have reported rates of 21%-47% (http://nacp.gov.pk). Furthermore, in the present study, the duration of selling sex and number of clients had an inverse relationship with condom usage. This is in concordance with a previous study conducted in Vietnam, where having five or less clients per day was associated with consistent condom use [20]. Inaccurate self-reporting of risky behaviours is the primary threat to the validity and utility of research among most-at-risk populations. The self-reported risk behaviours are often not correct and hence have implications for designing future strategies and interventions [23–25].

In our study, 91% of the women selling sex were married. To interpret the results, it is important to understand the context of the decision to sell sex as a married woman in Pakistan. Two thirds of Pakistan's population live on less than Euro 1.34 per day [5], with a total fertility rate of 4.1, an inflation rate of 7.7% (including a food inflation rate of 10%), and a high unemployment rate of 12.6%(http://finance.gov.pk). As a result, the struggle for survival has forced a number of married women, some of whom participated in our study, to sell sex. Approximately 5% of our study participants were forced to sell sex by their mother-in-law. This contradicts reports by both Saeed and Brown, which says that girls marrying into a family of women selling sex in Lahore are commonly forbidden to sell sex and kept in separate quarters at home [9, 26]. In our study however, the majority of women did not belong to families of women selling sex.

About 7% of our participants had female clients. This is a new finding and calls for further study of what appears to be a new type of social behaviour in Pakistan. To our knowledge, similar findings have not been reported before.

The low level of knowledge about HIV and AIDS in the present study is consistent with the results of the NACP survey conducted in 12 major cities of Pakistan (http://nacp.gov.pk). However, most alarming is that half of the participants did not have the correct knowledge about HIV transmission and prevention. Fifty five percent claimed that they were not vulnerable to AIDS. This might increase their likelihood of engaging in behaviours that make them vulnerable to exposure to HIV/STIs. Although the perception of being safe from HIV infection and AIDS is lower than the 82% reported by Bokhari [16], this is a huge challenge to those working with HIV/AIDS information among women selling sex in Pakistan.

Both HIV infection and STIs among women selling sex were relatively low in our study, which is possibly due to a high condom use, relatively low numbers of sexual partners and availability of clinical services, including syndromic management. However, there exists a high risk for a concentrated HIV epidemic among women selling sex due to their low level of knowledge about HIV, risk behaviour and sexual practices.