Eluwa, George I. MBBS, MS*; Strathdee, Steffanie A. PhD†; Adebajo, Sylvia B. MBBS, MPH, MS*; Ahonsi, Babatunde PhD*; Azeez, Aderemi MBBS, MPH‡; Anyanti, Jennifer MBBS, MPH§
Nigeria, the most populous country in Africa with an estimated population size of 167 million people in 2011,1 has approximately 3 million people living with HIV.2 The HIV epidemic in Nigeria is classified as generalized because two-thirds of new infections occur among persons who are regarded as “low risk” (ie, do not report engaging in high-risk sex).3 Key target groups comprising female sex workers (FSWs), injecting drug users, and men who have sex with men together constitute only 1% of the Nigerian population yet account for about a third of new HIV infections, with FSWs directly accounting for about 20%.3
HIV prevalence among FSWs is heterogeneous across Nigeria, ranging from a high of about 50% in Kano state and the Federal Capital Territory to a low of 24% in Lagos state, among brothel-based FSW in 2007.4 Studies have shown that targeting most-at-risk groups is a more cost-effective strategy for reducing population level HIV incidence and prevalence, irrespective of whether the type of HIV epidemic is concentrated or generalized.5–8 Characterizing the HIV epidemic among FSWs at national and regional level is especially important because the number of HIV-positive FSWs is the best predictor of country-level HIV prevalence.9
In Nigeria, sex work is illegal only for male sex workers, which is termed “an offence against morality.”10 Laws pertaining to female sex work include aiding and abetting prostitution, living off the earnings of a prostitute, and managing a brothel, but the act of prostitution itself is not illegal.10 Various prevention programs have been implemented for FSWs by both government and nongovernmental organizations in Nigeria. These include programs designed to increase their knowledge of HIV, increase condom use with clients through provision of free condoms and condom negotiation skills, free HIV voluntary counseling and testing (VCT) through venue-based outreach and linkages to dedicated VCT centers, application of peer education plus prevention activities for brothel-based FSWs (BB-FSWs) and Priority for Local AIDS Control Efforts, an intervention outreach program for non–brothel-based FSWs (NBB-FSWs).11,12 These activities have been implemented since 2007 and are usually implemented by trained peer educators who may or may not be FSWs.
There have been 2 successful rounds of Integrated Biological and Behavioral Surveillance Surveys (IBBSSs) in 2007 and 2010 in Nigeria. This article compares HIV prevalence overall and at the regional level between 2007 and 2010 and characterizes correlates of HIV among FSWs in Nigeria. We were specifically interested in determining whether HIV prevalence among FSWs had changed between 2007 and 2010 and whether FSWs in specific states or venues (ie, BB-FSW) were more or less likely to be HIV-infected. Such information is useful for informing, planning, and providing targeted programming for HIV prevention among FSWs in Nigeria.
The 2007 IBBSS was conducted in 6 states, whereas in 2010, 9 states were surveyed. To allow for comparison, only the 6 Nigerian states that were included in both rounds were analyzed in this article; Anambra (south east), Cross River and Edo (south south), Federal Capital Territory (north central), Kano (north west), and Lagos (south west). These states are located in 5 of the 6 geopolitical zones in Nigeria and thus adequately present a national picture of the burden of HIV among FSWs in Nigeria.
Study Population and Sampling Design
In the IBBSS, an FSW was defined as any biological female aged 15 years and older who receives money or other gifts/incentives in exchange for sex in areas such as brothels, bars, restaurants, nightclubs, hotels, or on the street.4 Respondents self-identified as FSWs at targeted hot spots and brothels. Furthermore, questions were asked about duration of sex work, reasons for being in sex work, and if sex work was full or part time to further ascertain that the respondents were FSWs. Taking into account the venues where sexual transactions took place most frequently (ie, brothels, bars, restaurants, nightclubs, hotels, on the street, or from a residence), they were classified as brothel-based FSW (BB-FSW) and non–brothel-based FSW (NBB-FSW). Key informants, mainly FSWs, members of the community, and nongovernmental organizations working with sex workers were consulted before the commencement of the survey.
Because of differences in FSW profiles, a mixed sampling method was used to recruit BB-FSW and NBB-FSW. BB-FSWs were recruited using 2-stage cluster sampling with brothels being the cluster unit. Lists of brothels in selected states used for the 2007 IBBSS were obtained from the State AIDS Control Program and updated for the 2010 IBBSS. The approximate number of FSWs in each brothel was obtained from key informant interviews with the brothel managers. With probability proportional to size of FSW per cluster, a fixed number of FSWs were then sampled from each cluster. NBB-FSW were recruited using time location sampling. A list of streets, bars, nightclubs, and other places where FSWs usually meet was generated in each state. The peak days and times were identified and used to define time location sampling clusters. NBB-FSWs were selected using probability proportional to size with a fixed number of sex workers recruited from each cluster. When the estimated number of NBB-FSW in the state was less than the desired sample size, a “take-all” approach was used. These sampling methods were used in both rounds of IBBSS.
A total sample size of 250 each of BB-FSW and NBB-FSW per state was required using a design effect of 1.5 and a level of precision of 0.05 to detect a 15% change in behavioral and biological components between the 2007 and 2010 IBBSS. Because reported condom use with clients in the 2007 IBBSS was high (ie, 93% overall), the behavioral component used was “consistent condom use with boyfriends in last 12 months.” Written informed consent was obtained from all participants for both surveys. Ethical approval was obtained from the Protection of Human Subjects committee of Family Health International, North Carolina; the US Centers for Disease Prevention and Control; and the National Institute for Medical Research, Nigeria in the 2007 IBBSS, whereas for the 2010 IBBSS, in addition to the aforementioned, ethical approval was obtained from Population Council, New York.
Behavioral data were obtained using a structured precoded questionnaire that was administered by trained interviewers in English and pidgin English (local form of English). The questionnaire captured data on sociodemographic characteristics, sexual risk behaviors including number of clients in last week of sex work, condom use by different partner types at last sexual contact, frequency of condom use over the last 12 months, charging more money for sex without condoms and comprehensive knowledge of HIV. A boyfriend was defined as a male partner who was engaged in a sexual and emotional noncommercial-based relationship with an FSW while a regular partner was defined as a spouse or live-in partner. Comprehensive knowledge was assessed by providing correct answers to 5 questions related to HIV as defined by United Nations General Assembly Special Session on HIV/AIDS.13 Nonmonetary incentives were provided, such as wristbands, T-shirts, and condoms valued at less than a dollar.
Rapid test using blood samples obtained from a finger prick was used to conduct HIV test using the parallel algorithm with Determine (Alere Medical, USA) and Unigold (Trinity Biotech, Plc, Bray, Ireland) simultaneously in accordance with Nigeria's national HIV testing protocol. All discordant tests were retested using Stat Pak (Chembio Diagnostic Systems, New York). Using linked anonymous methods to correlate HIV status with sexual risk behaviors, blood samples collected were coded by the HIV counselling and testing personnel with the same code as the questionnaire. All participants received pre- and posttest counseling, and respondents who tested positive were referred to the nearest comprehensive HIV care and treatment sites for further management.
Analyses included descriptive statistics of demographic, behavioural, and biological variables. To measure significance of changes between 2007 and 2010 IBBSS, the Wilcoxon rank sum test was used to compare changes in continuous variables, whereas the χ2 test was used to test statistical significance between categorical variables. Analyses were disaggregated by type of FSW (ie, BB-FSW vs. NBB-FSW), year of IBBSS (2007 vs. 2010), and across the 6 states. Bivariate logistic regression analysis was used to test associations between HIV and predictor variables. Variables significant at P ≤ 0.2 were considered for inclusion in multivariate logistic regression models to identify factors independently associated with HIV prevalence among FSW while controlling for potential confounders. Variables attaining significance at a P value ≤ 0.05 in the multivariate analysis were retained, based on the likelihood ratio test.
General Characteristics of FSW in 2007 and 2010 IBBSS
Table 1 shows general characteristics of respondents. The number of study participants was 2971 in 2007 and 2963 in 2010. Overall, FSW in 2010 were slightly older than those in 2007 with median age of 26 years [interquartile range (IQR), 23–29] versus 25 years (IQR, 22–29), respectively (P < 0.001). A significantly higher percentage of FSWs had completed secondary level education in 2010 (50%) compared with that in 2007 (42%, P < 0.001), whereas about a third of FSWs had ever been married in 2007 and 2010 (P = 0.23). Consistent condom use with boyfriends and clients were lower among FSWs in 2010 when compared to 2007; 23% vs. 25% (P = 0.02) and 90% vs. 93% (P < 0.001), respectively. A higher proportion of BB-FSW and NBB-FSW had completed secondary level education in 2010 compared to respondents in 2007: 40% versus 30%; P < 0.001 and 61% versus 55%; P = 0.003, respectively. Higher proportions of NBB-FSW reported ever being married in 2010 compared to 2007 (31% vs. 25%; P < 0.001), whereas among BB-FSW, there was no difference (36% vs. 37%; P = 0.19).
Comprehensive Knowledge of HIV and Exposure to HIV Prevention Interventions
Overall, there was a slight but significant increase in proportion of FSWs with comprehensive knowledge about HIV/AIDS between 2007 and 2010 (33% vs. 39%; P < 0.001). This was also observed among BB-FSW group (26% vs. 38%; P < 0.001); however, among NBB-FSWs, comprehensive knowledge remained the same at 40% in 2010 and 2007. Higher proportions of respondents reported having VCT the year before the survey in 2010 than in 2007 (72% vs.61%; P < 0.001), which was consistent both in BB-FSW (75% vs. 65%; P < 0.001) and in NBB-FSW (68% vs. 54%; P < 0.001). Proportion of NBB-FSWs and BB-FSWs reached by a peer educator in 2010 was less than that achieved in 2007, 45% versus 49%; P < 0.001 and 63% versus 67%; P = 0.05, respectively.
Sexual Risk Behavior Among FSWs
The median number of clients in the week preceding the survey was not significantly different in between 2007 and 2010 (10 vs. 11; P = 0.57). However, there was a 50% increase (from 16 to 25 clients per week) in the median number of clients in last week among BB-FSW (P < 0.001) and about a 30% decrease (from 8 to 6 clients per week) among NBB-FSW (P < 0.001) between 2007 and 2010, respectively. Overall there was a decline in 2010 compared to 2007 for consistent condom use with clients in the month before the survey (90% vs. 93%; P < 0.001) and among NBB-FSWs (84% vs. 93%; P < 0.001), whereas the proportion remained the same for BB-FSW (95% vs. 94%; P = 0.42). Consistent condom use with boyfriends in the last 12 months before survey was very low in both 2007 and 2010 [25% vs. 23% (P = 0.02)]. When comparing condom use with boyfriends, among BB-FSW, it was lower in 2010 compared to 2007 (18% vs. 23%; P < 0.001), whereas among NBB-FSWs, there was no significant difference between 2007 and 2010 (27% vs. 30%; P = 0.08). The proportion of respondents who reported receiving more money for sex without condom also increased in 2010 overall (7% vs. 4%; P = 0.002) and among NBB-FSW (9% vs. 4%; P < 0.001). For BB-FSWs, there was, however, a marginal but significant decrease in 2010 compared to 2007 (2% vs. 4%; P = 0.01).
Change in HIV Prevalence
As seen in Table 1, overall HIV prevalence among FSWs declined by more than one-third between 2007 and 2010 (32% vs. 20%; P < 0.001) with a larger drop observed among NBB-FSW (28% vs. 16%; P < 0.001). For BB-FSWs, HIV prevalence declined by more than one-third between 2007 and 2010 (37% vs. 23%; P < 0.001). Respondents aged 25 to 34 years had the highest burden both in 2007 and 2010 (47% vs. 56%; P = 0.02). Compared with those who had completed secondary level education, respondents who had not done so had higher HIV prevalence in 2007 (37% vs. 26%, P < 0.001) and 2010 (23% vs. 17%, P < 0.001) (data not shown).
Change in HIV Prevalence, HIV Knowledge, and Risk Behaviors by State
There were significant declines in HIV prevalence in all states surveyed (Table 2) with Edo state yielding the largest decline between 2007 and 2010 (30% vs. 17%; P < 0.0001). HIV prevalence declined by more than one-third from 2007 to 2010 in Anambra (28% vs. 18%; P = 0.05), Cross River (22% vs. 14%; P < 0.001), and Lagos state (20% vs. 13%; P < 0.001), whereas it declined one-quarter in the Federal Capital Territory (FCT) (39% vs. 29%; P < 0.001). The lowest decline was 18.4% in Kano state (38% vs. 31%, P < 0.001).
Anambra and Lagos states both had significant increases in the proportion of respondents who had comprehensive knowledge of HIV between 2007 and 2010: 19% versus 58%; P < 0.001 and 26% versus 38%; P < 0.001, respectively, whereas Edo state and the FCT had significant declines in the proportion of respondents who had comprehensive knowledge of HIV: 32% versus 20%; P < 0.001 and 40% versus 33%; P = 0.01, respectively.
Only FSWs in Anambra state showed an increase in consistent use of condom with clients in the month before the survey in 2010 compared to 2007 (89% vs. 93%; P = 0.05). Cross River, Edo and Lagos states showed significant declines in consistent condom use with clients between 2007 and 2010 (Table 2). Although consistent condom use with boyfriends was less than a third across all states except Lagos state, only Anambra state had a significant increase between 2007 and 2010 (15% vs. 27%; P < 0.001).
Factors Associated With HIV Infection
Table 3 outlines results from the multivariate analysis logistic regression that was used to determine factors independently associated with HIV infection. Compared to respondents aged 15–24 years, those aged 25–34 years were more likely to be HIV-positive [adjusted odds ratio (AOR): 1.45; 95% confidence interval (CI): 1.22 to 1.75]. Compared to Anambra state, FSWs located in the FCT (AOR: 2.03; 95% CI: 1.43 to 2.89) and Kano state (AOR: 2.66; 95% CI: 1.87 to 3.76) were more likely to be HIV-positive, whereas FSWs located in Lagos state (AOR: 0.61; 95% CI: 0.40 to 0.92) were less likely to be HIV-positive. BB-FSWs were also more likely to be HIV-positive (AOR: 1.37; 95% CI: 1.14 to 1.65) compared to NBB-FSW. Respondents in 2010 were less likely to be HIV-positive than respondents in 2007 (AOR: 0.79; 95%CI: 0.75 to 0.85), whereas respondents who had completed secondary level education were also less likely to be HIV-positive (AOR: 0.64; 95% CI: 0.53 to 0.77).
This study examining changes in the prevalence and correlates of HIV infection among FSWs in Nigeria identified several important observations. First, HIV surveillance data suggests that between 2007 and 2010, HIV prevalence declined by at least one-third in 4 of the 6 states studied, whereas HIV prevalence in FCT declined by one quarter. Second, HIV prevalence declined significantly among NBB-FSWs and BB-FSWs, with greater declines among the former subgroup. Third, although proportion of FSWs who reported consistent condom use with clients was high in all 6 states in 2007, 3 states recorded lower levels of condom use with clients in 2010, and among NBB-FSWs and BB-FSWs in 2 of the 6 states, the proportion of FSWs who reported consistently using condoms with boyfriends significantly declined. Finally, factors independently associated with HIV infection among Nigerian FSWs included older age, being interviewed in 2010 versus 2007, residence in FCT and Kano state, being a BB-FSW, and not having completed secondary school. These findings have salient implications for HIV prevention planning in Nigeria.
These findings suggest that Nigeria is making significant progress in reducing the burden of HIV among FSW with an overall 39% decrease in the 6 states surveyed between 2007 and 2010. Furthermore, at the state level that represents a diverse geographic distribution and considerable sociocultural and socioeconomic heterogeneity, there were also significant declines in HIV prevalence, ranging from 26% to 45% between 2007 and 2010, which is consistent with the high levels of consistent condom use with clients reported both in 2007 and 2010, and increases in the level of comprehensive HIV knowledge, and uptake of VCT. Together, these findings suggest that HIV preventions programming may have begun to have an impact on HIV prevalence in these FSW populations. FSWs located in FCT and Kano states were more likely to be HIV-positive and given that both states had the highest HIV prevalence in both 2007 and 2010, more research is needed to further understand factors driving the epidemic in these states.
Brothel-based workers were also more likely to be HIV-positive than NBB-FSW, probably because the latter is comprised of a heterogeneous subgroup of women, which includes escorts and bar-based FSWs. The higher client load among BB-FSWs and their tendency to be less educated and be from lower socioeconomic backgrounds may help explain this finding.
Clearly, more focused subgroup programming may be needed to reduce the risk among Nigerian BB-FSW. In Thailand and the Dominican Republic, structural interventions such as 100% condom campaigns, offered alone or in combination with efforts to strengthen community solidarity, have significantly reduced HIV/Sexually Transmitted Infections (STI) incidence within this subpopulation.14,15 Promoting community solidarity involves regular workshops and follow-up meetings with sex workers, establishment owners and managers, and other establishment employees. The purpose of these encounters is to encourage and strengthen a sense of solidarity and collective commitment toward HIV and STI prevention, and discussions include the role that each actor could play in supporting sex workers to use condoms with their partners. Such approaches should be considered in Nigeria and have been used in some instances in Nigeria, albeit on a small scale. Before Nigeria lauds itself for progress, we propose a caveat. The lowest HIV prevalence recorded (14%) is still 3 times higher than the national prevalence of 4.6%2 and thus gives an insight into the magnitude of work still remaining to reduce the burden of HV among FSW. The appreciable decline observed across all states reinforces the urgent need for state-specific evidenced-based HIV prevention programming for FSWs to further reduce the burden of HIV among FSW.
Sexual Risk Behavior
A number of studies have reported weak correlations between behavioral changes and HIV risk.5,16–18 Given the complex epidemiology of HIV and the various risk factors associated with HIV, what is the best indicator to measure risk of HIV transmission? Pinkerton et al19 propose that the total number of unprotected sex acts is the best marker of risk for less infectious STI, such as HIV. Across the 6 Nigerian states surveyed, consistent condom use with clients was high in 2007 but significantly declined in Cross River, Edo and Lagos by 2010. Concerted efforts must be made to immediately address this trend before the situation significantly worsens. The proportion of respondents who reported charging more money for sex without condoms also increased between 2007 and 2010 and suggests that reducing high-risk behavior would be futile without parallel programs focused on clients.
Condom use between FSWs and their boyfriends was low across all states surveyed. This finding has been observed globally20–24 and shows that the risk of HIV transmission among FSWs will continue to remain high as long as consistent condom use remains low with noncommercial partners. In Kenya, Voeten et al25 reported that <40% of FSWs used condoms with their boyfriends, whereas in Glasgow, only 17% of FSWs used condom consistently with noncommercial partners.26 Of concern was the finding that the proportion of FSWs reporting consistent condom use with boyfriends significantly declined in 2 of the 6 states surveyed. Because there is evidence that FSWs' intimate male partners engage in high-risk sexual and drug use behaviors in other settings,24,25,27 these men may represent a potential transmission bridge between the general population and FSWs. Further research to identify factors associated with consistent use of condom is needed to inform programming and practice. Efforts to track the number of unprotected sex acts by partner type must be encouraged and should inform future surveys.
This study compared 2 cross-sectional surveys and thus must be interpreted with caution as no prospective control group exists for comparison. Because we did not track coverage of HIV prevention programs in the states surveyed, the possibility exists that observed behavioral changes may have occurred independently of the HIV prevention programs in the states. However, given the large sample size, the geographical diversity of respondents surveyed, and the varied but consistent decline in HIV prevalence across all states surveyed, it is unlikely that the decline in HIV prevalence occurred by chance. Furthermore, mortality, mobility, and other factors may have contributed to this decline as well. Another limitation of the study is that information was self-reported and reports of condom use with clients may be subject to social desirability bias. Given the observed decline in HIV prevalence, and the reports of lower condom use with boyfriends, it is unlikely that condom use is seriously over-reported. Further research using methods designed to reduce social desirability bias, such as audio computer-assisted self-interview may provide more insight into sexual risk behaviors among FSW. Additional studies are also needed to examine the influence of drug and alcohol use on FSWs' risk for HIV infection, especially since a recent study reported high HIV prevalence among female injecting drug users in some Nigerian states.28
This is the first study measuring changes in HIV prevalence and sexual risk behaviors between 2 rounds of IBBSS in Nigeria. There seems to be significant positive impacts of HIV prevention programming among FSW with regard to reduced HIV prevalence and sustained condom use with clients. However, other behavioral indicators of HIV/STI risk such as consistent condom use with boyfriends are far below the desired level. Programming must be diversified to address drivers of the epidemic in BB-FSW and NBB-FSW. Possible interventions include expansion of the Priority for Local AIDS Control Efforts intervention for NBB-FSW, peer education plus events for BB-FSW, and 100% condom programs with community-based approaches that may include community solidarity, environmental cues that serve as visible reminders, accessible clinical services, good program monitoring to improve ART adherence, and supportive policies and regulations. Evidence-based strategies are needed to increase condom use with boyfriends and regular partners because these are potential bridges to the general population. There is an urgent need to reinforce successful strategies and scale up HIV prevention services to sustain the decline in HIV prevalence among FSWs in Nigeria.
The authors wish to acknowledge the HIV and AIDS division, Federal Ministry of Health, Nigeria for successfully conducting the IBBSS in collaboration with Family Health International 360, the Enhancing Nigeria's Response to HIV and AIDS Program, Society for Family Health, and Population Council. The authors also thank all members of the Technical Committee, field supervisors, laboratory scientists, interviewers, and all those who contributed to the success of the IBBSS.
1. National Population Commission. Nigeria's over 167 million population: implications and challenges. Available at: http://www.population.gov.ng
. Accessed March 23, 2012.
2. Federal Ministry of Health, Nigeria. National HIV Sero-Prevalence Sentinel Survey Among Pregnant Women Attending Antenatal Clinics in Nigeria; 2010. Technical report. HIV/AIDS Division, Department of Public Health, Abuja, Nigeria.
3. National Agency for the Control of AIDS. Modes of HIV Transmission in Nigeria. Abuja, Nigera: NACA: 2010.
4. National AIDS/STI Control Programme, Federal Ministry of Health. HIV/STI Integrated Biological Behavioural Surveillance Survey (IBBSS). Abuja, Nigeria: Federal Ministry of Health; 2007.
5. Luchters S, Chersich MF, Rinyiru A, et al.. Impact of five years of peer-mediated interventions on sexual behaviorr and sexually transmitted infections among female sex workers in Mombasa, Kenya. BMC Public Health. 2008;8:143.
6. WHO, UNAIDS. Progress on Global Access to HIV Antiretroviral Therapy: A Report on “3 by 5” and Beyond. Geneva: WHO; 2006.
7. Boily MC, Lowndes C, Alay M. The impact of HIV epidemic phases on the effectiveness of core group interventions: insights from mathematical models. Sex Transm Infect. 2002;78:i78–i90.
8. Rekart ML. Sex-work harm reduction. Lancet. 2005;366:2123–2134.
9. Talbott JR. Size matters: the number of prostitutes and the global HIV/AIDS pandemic. PLoS One. 2007;2:e543.
11. Society for Family Health, Nigeria. An Evaluation of a Quasi-Experimental Community Level HIV Intervention in Nigeria. Abuja, Nigeria: Society for Family Health; 2005.
12. Society for Family Health Nigeria. Assessing Behavior Change among HIV Risk Groups in Nigeria: Indirect Interventions through Civil Society Organizations. Abuja, Nigeria: Society for Family Health; 2009.
13. Joint United Nations Programme on HIV/AIDS (UNAIDS). Monitoring the Declaration of Commitment on HIV/AIDS: Guidelines on Construction of Core Indicators; 2010:54. 2010 reporting.
14. Rojanapithayakorn W, Hannenberg R. The 100% condom program in Thailand. AIDS. 1996;10:1–7.
15. Kerrigan D, Moreno L, Rosario S, et al.. Environmental-structural interventions to reduce HIV/STI risk among female sex workers in the Dominican Republic. Am J Public Health. 2006;96:120–125.
16. Wi T, Ramos ER, Steen R, et al.. STI declines among sex workers and clients following outreach, one time presumptive treatment, and regular screening of sex workers in the Philippines. Sex Transm Infect. 2006;82:386–391.
17. Thomsen SC, Ombidi W, Toroitich-Ruto C, et al.. A prospective study assessing the effects of introducing the female condom in a sex worker population in Mombasa, Kenya. Sex Transm Infect. 2006;82:397–402. Abstract.
18. Celentano D. Decreasing incidence of HIV and sexually transmitted diseases in young Thai men: evidence for success of HIV/AIDS control and prevention program. AIDS. 1998;12:29–36.
19. Pinkerton SD, Chesson HW, Layde PM. Utility of behavioural changes as markers of sexually transmitted disease risk reduction in sexually transmitted disease/HIV prevention trials. J Acquir Immune Defic Syndr. 2002;31:71–79.
20. Rosenthal D, Oanha TT. Listening to female sex workers in Vietnam: influences on safe-sex practices with clients and partners. Sex Health. 2006;1:21–32.
21. Mills S, Benjarattanaporn P, Bennett A, et al.. HIV risk behavioral surveillance in Bangkok, Thailand: sexual behavior trends among eight population groups. AIDS. 1997;11:S43–S51.
22. Philpot CR, Harcourt CL, Edwards JM. A survey of female prostitutes at risk of HIV infection and other sexually transmissible diseases. Genitourin Med. 1991;5:384–388.
23. Jamner MS, Wolitski RJ, Corby NH, et al.. Using the theory of planned behavior to predict intention to use condoms among female sex workers. Psychol Health. 1998;2:187–205.
24. Spina M, Mancuso S, Sinicco A, et al.. Human immunodeficiency virus seroprevalence and condom use among female sex workers in Italy. Sex Transm Dis. 1998;25:451–454.
25. Voeten HA, Egesah OB, Varkevisser CM, et al.. Female sex workers and unsafe sex in urban and rural Nyanza, Kenya: regular partners may contribute more to HIV transmission than clients. Trop Med Int Health. 2007;12:174–182.
26. Green ST, Goldberg DJ, Christie PR, et al.. Female streetworker—prostitutes in Glasgow: a descriptive study of their lifestyle. AIDS Care. 1993;5:321–335.
27. Day S, Ward H, Perrotta L. Prostitution and risk of HIV: male partners of female prostitutes. BMJ. 1993;307:359–361.
28. Eluwa GI, Strathdee SA, Adebayo SB, et al.. A profile on HIV prevalence and risk behaviors among injecting drug users in Nigeria: should we be alarmed? Drug Alcohol Depend. 2012. In press.
© 2012 Lippincott Williams & Wilkins, Inc.