Share this article on:

Identifying At-Risk Populations in Kenya and South Africa: HIV Incidence in Cohorts of Men Who Report Sex With Men, Sex Workers, and Youth

Price, Matthew A. MSPH, PhD*; Rida, Wasima PhD; Mwangome, Mary MBChB, MSc; Mutua, Gaudensia MBChB, MPH§; Middelkoop, Keren MBChB, PhD||; Roux, Surita MBChB, MPH; Okuku, Haile S. MSc; Bekker, Linda-Gail MBChB, FCP, PhD; Anzala, Omu MBChB, PhD§; Ngugi, Elizabeth PhD§; Stevens, Gwynn PhD*; Chetty, Paramesh ND*; Amornkul, Pauli N. MD, MPH*; Sanders, Eduard J. MD, PhD‡,¶

JAIDS Journal of Acquired Immune Deficiency Syndromes: February 1st, 2012 - Volume 59 - Issue 2 - p 185–193
doi: 10.1097/QAI.0b013e31823d8693
Epidemiology and Prevention

Objective: To identify and describe populations at risk for HIV in 3 clinical research centers in Kenya and South Africa.

Design: Prospective cohort study.

Methods: Volunteers reporting recent sexual activity, multiple partners, transactional sex, sex with an HIV-positive partner, or, if male, sex with men (MSM; in Kenya only) were enrolled. Sexually active minors were enrolled in South Africa only. Risk behavior, HIV testing, and clinical data were obtained at follow-up visits.

Results: From 2005 to 2008, 3023 volunteers were screened, 2113 enrolled, and 1834 contributed data on HIV incidence. MSM had the highest HIV incidence rate of 6.8 cases per 100 person-years [95% confidence interval (CI): 4.9 to 9.2] followed by women in Kilifi and Cape Town (2.7 cases per 100 person-years, 95% CI: 1.7 to 4.2). No seroconversions were observed in Nairobi women or men in Nairobi or Cape Town who were not MSM. In 327 MSM, predictors of HIV acquisition included report of genital ulcer (Hazard Ratio: 4.5, 95% CI: 1.7 to 11.6), not completing secondary school education (HR: 3.4, 95% CI: 1.6 to 7.2) and reporting receptive anal intercourse (HR: 8.2, 95% CI: 2.7 to 25.0). Paying for sex was inversely associated with HIV infection (HR: 0.2, 95% CI: 0.04 to 0.8). 279 (13.0%) volunteers did not return after the first visit; subsequent attrition rates ranged from 10.4 to 21.8 volunteers per 100 person-years across clinical research centers.

Conclusions: Finding, enrolling, and retaining risk populations for HIV prevention trials is challenging in Africa. African MSM are not frequently engaged for research, have high HIV incidence, need urgent risk reduction counseling, and may represent a suitable population for future HIV prevention trials.

Supplemental Digital Content is Available in The Text.

*Department of Medical Affairs, International AIDS Vaccine Initiative, New York, NY

Biostatistics consultant, Arlington, VA

Centre for Geographic Medicine-Coast, Kenya Medical Research Institute, Kilifi, Kenya

§Kenya AIDS Vaccine Initiative, Nairobi, Kenya

Department of Medicine, Institute of Infectious Diseases and Molecular Medicine, Desmond Tutu HIV Centre, University of Cape Town, Cape Town, South Africa

Nuffield Department of Clinical Medicine, University of Oxford, Headington, United Kingdom

Correspondence to: Matthew A. Price, MSPH, PhD, 125 Broad Street, New York, NY 10004 (e-mail:

Supported in part by the generous support of the American people through the United States Agency for International Development.

Presented in part at The 18th Conference on Retroviruses and Opportunistic Infections, 2011, Boston, MA.

The contents are the responsibility of the study authors and do not necessarily reflect the views of United States Agency for International Development or the United States Government.

Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal's Web site (

This article is published with permission of the Director of KEMRI.

The authors have no conflicts of interest to disclose.

Received June 2, 2011

Accepted October 14, 2011

Back to Top | Article Outline


Over two-thirds of all persons living with HIV and 70% of all new HIV infections are estimated to be in sub-Saharan Africa, often described as a generalized epidemic.1 An increasing number of HIV prevention clinical trials have been conducted at African research institutions,2,3 focusing on new prevention technologies from circumcision4–6 to topical microbicides,7,8 antiretroviral therapy as prevention,9 and vaccines.3,10,11

To conduct large-scale HIV vaccine and other intervention trials, successfully identifying, enrolling, and retaining persons at high risk for HIV infection can improve efficiency. For cost and logistics considerations, an HIV incidence of at least 2–3 cases per 100 person-years has been recommended.12–14 Although higher incidence can translate into a smaller sample size, enrollment of high-risk populations must include HIV prevention counseling to modify behavior. This, in addition to a standard of prevention that increasingly will include more prevention technologies,9,15,16 may decrease HIV incidence further8,17,18 and reduced power to detect a significant effect has plagued previous prevention trials.19,20 High rates of pregnancy in marginalized high-risk populations can lead to high trial dropout rates unless effective retention strategies are used including user-independent contraceptive methods.7 In addition, marginalized populations, often at increased risk for HIV, may also be at higher risk for attrition due to migration. These factors all contribute to the complexity of trial design, participant recruitment, and volunteer retention.

The International AIDS Vaccine Initiative supports a network of clinical research centers (CRCs) in sub-Saharan Africa that conduct cohort studies in at-risk populations to prepare for HIV clinical trials by generating data on risk behavior, HIV incidence, volunteer retention, and pregnancy rates. HIV discordant couples have been of special focus for several CRCs since 2004.21,22 Since 2005, to identify potential and alternative source populations for future trials, 3 CRCs started prospective cohorts focused on other at-risk groups, including male and female sex workers (SWs), younger participants, and men who report sex with men (MSM). Here we present our results working with these populations to inform future HIV intervention trials.

Back to Top | Article Outline


Cohort Descriptions

A prospective cohort study conducted at 3 diverse CRCs in Kenya and South Africa has been supported by the International AIDS Vaccine Initiative. The sites included the densely populated areas of the Kangemi district in Nairobi, Kenya; the Kilifi district north of Mombasa, Kenya; and in Gugulethu in Cape Town, South Africa. Each CRC defined eligibility criteria with the goal of recruiting volunteers at risk for HIV infection. Recruitment of potential volunteers involved outreach to voluntary counseling and testing (VCT) centers, “hotspots” such as bars and brothels, and peer recruitment (ie, “snowball” recruitment) of harder-to-reach populations such as SWs or MSM (Tables 1, 2). Peer recruitment among the general population was employed in South Africa in a community, where antenatal care HIV prevalence rates were 28% in 2005.23





Potential participants were told about the study and screened for eligibility which included a self-report of the following: (1) recent sexually transmitted infection (STI); (2) sex with an HIV-positive partner; (3) receiving goods or money for sex (SW); (4) soliciting sex for goods or money (client of SW); (5) inconsistent condom use; (6) engaging in any anal sex (both men and women, Kilifi only); (7) MSM behavior (Nairobi only); or (8) report of any sexual activity (Cape Town only). The lower age limit of volunteers was 16 years in Cape Town and 18 years in the 2 centers in Kenya (Table 1).

Back to Top | Article Outline

Study Procedures

After obtaining informed consent, demographic and risk behavior data were collected, and a standardized physical and genital examination was performed by a study clinician and circumcision status determined. Specimens were collected at screening for HIV and syphilis. Symptomatic STIs were syndromically treated according to country-specific guidelines.

Volunteers were given VCT that included risk-reduction counseling, condom provision, and lubricant provision (for MSM). HIV antibody rapid test (Table 1) results were given to the volunteer on the same day. When rapid test results were discordant, a third tie-breaker test was used (Table 1). Specimens were collected for p24 antigen ELISA (enzyme-linked immunosorbent assay) testing (Coulter p24 ELISA until April 2007, Vironostika HIV p24 ELISA thereafter) and stored at −80°C. Antigen test results were available within 1 week of the volunteer's visit.

Volunteers returned for follow-up every 3 months at the Kenyan CRCs and monthly in Cape Town. Follow-up was for 2–4 years in Kenya, where volunteers who remained at risk after 2 years were invited to remain on study, and for 1 year in Cape Town. Volunteers were defined as lost-to-follow-up if they discontinued study participation before the completion of the study for any reason, with the exception of incident HIV infection. Follow-up visits included VCT, risk behavior assessment with additional counseling, urine pregnancy tests for women, a symptom-directed physical examination, and both HIV antibody and p24 antigen testing (Table 1). If pregnant, women were retained in the study and referred to antenatal care. Incident HIV infection was defined as 2 positive rapid HIV antibody tests or a positive p24 antigen test that was confirmed by antibody seroconversion within 3 months. Blood collected at every visit was stored, and annual syphilis screening by RPR with confirmation by TPHA was performed. Volunteers were provided treatment on site for minor illnesses and referred for additional care when treatment was not available. Volunteers were compensated for their time and travel costs incurred from study participation.

Back to Top | Article Outline

Ethical Considerations

Ethics approval was obtained from the Kenya Medical Research Institute Ethics Committee, Kenyatta National Hospital Ethics and Research Committee at the University of Nairobi, and the University of Cape Town Research Ethics Committee, South Africa. Special dispensation for unassisted consent to enroll 16-year and 17-year olds was obtained in Cape Town.

Back to Top | Article Outline

Data Analysis

Incidence rates for HIV infection, loss-to-follow-up, and clinically significant events (eg, pregnancy, syphilis, and report of STI ) were calculated as number of events per 100 person-years of observation and compared across CRCs and risk groups. Cox proportional hazards models were used to assess the association between time to HIV diagnosis and covariates including baseline demographic variables and time-varying measures of risk behavior. Four risk groups for the Cox models were identified as follows: MSM, non-MSM men, SW women, and non-SW women. The final model was stratified by CRC, and each risk group was analyzed separately. However, MSM was the only risk group with a sufficient number of events for multivariate analysis, and the results of Cox regression analyses are reported for this group only. Covariates associated with HIV diagnosis at a significance level of P <0.20 in univariate models were considered for inclusion in multivariate models. Multivariate models were constrained to include no more than 1 covariate for every 10 HIV diagnoses to ensure stable estimates of hazard ratios. Data analysis was conducted using STATA version 11.1 (College Station, TX) and R (version 2.10.1,

Back to Top | Article Outline


Study Population Characteristics

From July 2005 to June 2008, 3023 potential volunteers were screened for risk of HIV infection, 2172 (72%) were enrolled, and 1834 (61%) were included in the analysis; 716 in Kilifi, 653 in Nairobi, and 465 in Cape Town (Fig. 1 and Table 2). In Kilifi, fewer women were enrolled than men (58% of women screened vs. 66% of men, P = 0.001), and enrollees tended to be younger than nonenrollees (median age 25 years vs. 27 years, respectively, P = 0.02). Enrollment rates by sex and age were equivalent in Nairobi and Cape Town. Volunteers enrolled but excluded from the analysis included the following: 279 who did not return after the enrollment visit; 53 who, after screening, reported no eligibility criteria risk behavior at enrollment or subsequent visits; and 6 persons with incident HIV infection at screening, who were either antibody discrepant or antibody negative and p24 antigen positive at enrollment (Fig. 1). Age and gender did not vary significantly at any CRC between volunteers included in the analysis and volunteers who were not included.



Back to Top | Article Outline

Pregnancy and STI Rates

The incidence of pregnancy (Table 3) was lower in Kilifi compared with Nairobi and Cape Town (11.3 vs. 17.9 and 13.7 of 100 woman-years, respectively, P < 0.001). Kilifi reported a higher incidence of symptomatic STIs than Nairobi or Cape Town (22.8 vs. 3.7 and 4.4 cases per 100 person-years, respectively, P < 0.001). Syphilis rates were relatively high and similar in Kilifi and Cape Town but significantly lower in Nairobi (P < 0.001).



Back to Top | Article Outline

Study Retention

Study retention was a challenge; 279 of 2172 (13%) volunteers did not return after the enrollment visit (Fig. 1), and an additional 440/1834 (24%) volunteers left the study prematurely (Table 3). Attrition rates varied significantly by CRC with loss-to-follow-up rates of 20.4 volunteers lost per 100 person-years in Kilifi, 10.4 in Nairobi, and 21.8 in Cape Town (Table 3, P < 0.001). In Nairobi, MSM were significantly more likely to drop out than other volunteers (21.1 vs. 8.9 volunteers lost per 100 person-years, respectively, P < 0.001), but MSM were less likely to drop out in Kilifi (16.0 vs. 23.0 volunteers lost per 100 person-years, P = 0.004). In Kilifi, women SWs were more likely to remain in the study than non–SW women (21.2 vs. 36.3 volunteers lost per 100 person-years, P = 0.01), however, in Nairobi, the attrition rates were similar between women SWs and non–SW women.

Back to Top | Article Outline

HIV Incidence

HIV incidence was highest in MSM and similar across the 2 Kenyan CRCs recruiting MSM (6.8 cases per 100 person-years; 95% CI: 5.0 to 9.3; Table 3). The HIV incidence rate was similar in women in Kilifi and Cape Town (2.7 cases per 100 person-years; 95% CI: 1.7 to 4.2) and significantly lower than HIV incidence observed in MSM (P < 0.001). HIV incidence was low to nil among women in Nairobi and non-MSM men at all CRCs (Table 3).

Back to Top | Article Outline

Predictors of Incident HIV in Kenyan MSM

To characterize the population with the greatest risk of HIV infection in these cohorts, we analyzed the 327 Kenyan MSM from Kilifi and Nairobi in a Cox proportional hazards regression analysis for predictors of incident HIV infection (Table 4). MSM who reported less education were at an increased risk for HIV; not finishing secondary school was associated with a nearly 3.5-fold increase in risk (P = 0.001). Compared with MSM who reported no anal sex in the previous 3 months, the risk for HIV was 8 times higher (P < 0.001) among MSM who reported only receptive anal sex and about 3.5 times higher (P = 0.03) among MSM who reported both insertive and receptive anal sex. MSM who reported no anal sex had a similar risk of HIV acquisition to men who reported only insertive anal sex. Of the 146 MSM who reported no anal sex at 497 visits, 121 reported at least 1 sexual partner in the previous 4 weeks, suggesting that some sex remained unreported or was not specifically asked about (eg, oral sex). MSM who report genital ulcers in the previous 3 months had a 4-fold increase in risk for acquiring HIV compared with those without genital ulcers (P = 0.002). MSM with no data on genital ulcers also had a higher risk than those without genital ulcers (P = 0.03, Table 4); visits for which data were not available were frequently unscheduled, interim visits where participants requested HIV testing due to a perceived risky experience and where STI data were not likely to have been recorded. MSM who reported paying for sex were significantly less likely to become HIV infected during follow-up (P = 0.02). Additional analysis details are shown in Appendix A (see Supplemental Digital Content 1,



Back to Top | Article Outline


We present data on screening, enrollment, retention, demographics, pregnancy, and HIV incidence rates in multiple African populations enrolled into an HIV prevention trial preparedness study. Additionally, we present some of the first HIV incidence data from African MSM, including a multivariate analysis of predictors of HIV acquisition in this cohort.

Back to Top | Article Outline

Different Populations and Different Recruitment Strategies

The objective of the study was to identify and describe risk populations for future HIV prevention trials. Some important lessons in optimizing recruitment strategies to identify at-risk populations were learnt. HIV incidence was frequently lower than expected at all 3 CRCs presenting challenges in identifying true at-risk populations within the generalized HIV epidemics.

The research team in Cape Town elected to enroll younger volunteers thought to be at risk for incident HIV infection by virtue of living in a community known to have high HIV prevalence among antenatal care attendees.23 The same investigators reported results in an earlier cohort study in an adjacent community in Cape Town where they observed a high HIV incidence among male and female volunteers as young as 16 years.24 Recruitment strategies in these 2 cohorts were different: in the earlier higher incidence study individuals attending VCT services, and other health-related appointments were referred to the cohort study, whereas in this cohort study, individuals were largely peer recruited. Subsequent clinical trials in this district in Cape Town have also demonstrated high HIV incidence,25 in particular among young women. The differences in HIV incidence we observed may have been due to our recruitment methods, the frequency of HIV testing, and the need for more risk-targeted eligibility criteria.

We also found that recruiting female SWs in Nairobi was not an efficient strategy to establish an at-risk cohort for future HIV prevention trials. There was no HIV incidence in more than 300 women who self-identified as SWs. Kimani et al26 has reported a reduction in HIV-1 incidence in Nairobi SW recently. It may be possible that women SWs are more consistent in using condoms during sex work, their clients were less likely to be HIV infected, their HIV-infected clients were more likely to be using ART or some combination of the above.

Although the initial target populations in Kenya focused on SW and persons who reported other higher risk heterosexual behavior, it quickly became apparent that MSM were willing to come forward and participate in HIV prevention research. Feedback from study volunteers and health care providers was critical in designing MSM-friendly recruitment and retention strategies and led to hiring counselors familiar with MSM sensitivities.27 The HIV incidence we observed in this previously unstudied risk group was the highest in this report.

Back to Top | Article Outline

Pregnancy and STI Rates

Despite lower than expected HIV rates, we found a high rate of pregnancy among female volunteers that ranged from 11.3 to 17.9 per 100 person-years of follow-up depending on CRC. These rates are similar to what has been reported elsewhere in sub-Saharan Africa in the context of risk reduction and condom-use counseling.7,28 In addition, STI rates are particularly high in the cohorts from Kilifi, but also high in Nairobi and Cape Town. These pregnancy and STI rates suggest that unprotected sex was certainly being practiced during study follow-up. Because this was a preparedness study, pregnancy was not a cause for study termination, but in prevention trials, pregnancy may require a participant to stop receiving the HIV prevention intervention being studied (eg, a vaccine or microbicide). To assure adequate power, clinical trial designs will need to incorporate steps to try to minimize pregnancy risk and/or sufficiently power studies to accommodate pregnancy rates.

Back to Top | Article Outline


Attrition was high in these cohorts as follows: 13% of volunteers who enrolled never returned for a second visit; and an additional 24% of volunteers left the study prematurely. Some volunteers may have seen screening and enrollment as an opportunity for a single VCT and/or health care visit. This highlights the importance of careful counseling before enrollment and supports the use of 1 or more screening visits to gauge volunteer commitment before participation in a prevention trial. The high attrition may reflect study fatigue because regular study visits with HIV testing could be burdensome but may also reflect hidden HIV incidence with participants participating in riskier activities electing to not return. Anticipating volunteers at higher risk for dropping out and proactively developing strategies to keep these persons onstudy should be a priority in prevention trials.

Back to Top | Article Outline


This study has several limitations. The 3 CRCs employed different recruitment strategies and identified largely different at-risk populations. As a result, analysis across CRCs was difficult. The lower incidence in the non-MSM risk groups precluded any multivariate analysis of HIV risk associations in these cohorts.

Because some measure of prescreening often occurred before the enrollment visit, the baseline values for pregnancy, STI, and HIV (Fig. 1 and Table 3) do not represent estimates of prevalence that can be generalized beyond the screening population. For example, because potential volunteers were only recruited from VCT clinics if their HIV test results were negative, the proportion of HIV-infected persons at study screening (Fig. 1) cannot be generalized as an estimate of HIV prevalence in other similar groups of at-risk persons.

Back to Top | Article Outline

New Target Populations for Prevention Trials

With relatively few exceptions,18 the focus of prevention trials in adult Africans has been largely on heterosexual transmission of HIV.9,17,29–31 However, new reports are identifying MSM as an important at-risk population to consider for prevention research, and the incidence we observed in MSM is among the highest HIV incidence reported from Africa.32 Recent publications have described the identification and recruitment of MSM in Kilifi,33 including initial interview data34 and factors associated with prevalent HIV.35 Numerous studies document a high HIV prevalence among African MSM,36–38 but we present among the first incidence data and predictors of HIV acquisition in African MSM. We were able to identify that in MSM, lower educational achievement, genital ulcers, and receptive anal sex were associated with HIV acquisition. Paying for sex was inversely associated with HIV acquisition. This latter, unexpected finding, may have been due to adoption of safer behavior when purchasing sex. MSM who purchased sex may have been more likely to use condoms if they felt their sexual partner may have been HIV infected. Anecdotally, MSM who report paying for sex frequently reported purchasing sex from female sellers and using condoms with women. Additionally, MSM reported they felt women to be higher risk partners because they perceived anal sex with a man as lower risk.39

The high HIV incidence we report may indicate that MSM are a suitable population for prevention trial participation. However, trial design must take into consideration the route of exposure. Interventions that are thought to act at the point of exposure may have different outcomes in MSM versus non-MSM populations. For example, the rectal epithelium is more delicate than that of the vagina, and a microbicide designed for vaginal use may therefore not be suitable for the rectum.40 Beyond biology, same-sex sexual relations remains illegal in many countries in Africa, with consequent lack of HIV services, prevention messages, and general awareness among health providers.41 A recent UNAIDS report highlights the deficiencies in addressing the HIV prevention, care, and treatment needs of MSM and reinforces the importance of closing this gap from both a human rights and public health perspective.42 Because African MSM face considerable societal, legal, and social hurdles, careful consideration of the counseling and clinical needs, follow-up schedule, and social support is vital to ensure continuing research participation. Beyrer et al43 have recognized that flexibility in the integration of these services with existing services is critical and have reviewed multiple models of service provision for at-risk populations including MSM, and an MSM-specific training manual for health care workers in Africa has recently been published.44 Additional research to understand and reduce the barriers to recruitment and retention of African MSM is warranted.

Back to Top | Article Outline


Although HIV incidence data in sub-Saharan African remain sparse,45 this study joins a growing body of literature contributing to the understanding of local HIV epidemics, identification and enrollment of persons at risk for HIV infection, and site preparedness in anticipation of future HIV prevention trials in Africa.7,31,45–49 We were able to successfully enroll volunteers who included MSM, SWs and their clients and persons reporting multiple partners, inconsistent condom use, recent STI, report of any anal sex, and sex with a known HIV-positive person. However, HIV incidence was lower than expected among non-MSM, and retention was often challenging. Our work shows that Kenyan MSM are at very high risk for HIV infection, need urgent risk reduction interventions, and may be a suitable cohort for future HIV prevention studies.

Back to Top | Article Outline


We thank all the study volunteers who participated in this research. We thank the study staff for their work in maintaining these large study cohorts. We thank Sarah Yates of International AIDS Vaccine Initiative, and Nick Moodley, and Busi Mpila of the Perinatal HIV Research Center in South Africa for study data management and Contract Laboratory Services in Johannesburg for laboratory network support.

Back to Top | Article Outline


1. Joint United Nations Programme on HIV/AIDS (UNAIDS). UNAIDS Report on the Global AIDS Epidemic. Geneva, Switzerland: UNAIDS; 2010.
2. Excler JL. AIDS vaccine efficacy trials: expand capacity and prioritize. 'Throughout Africa, Asia and Latin America state-of-the-art clinics and laboratories exist where, 4 years ago, there were none'. Expert Rev Vaccines. 2006;5:167–170.
3. Fast PE, Kaleebu P. HIV vaccines: current status worldwide and in Africa. AIDS. 2010;24(suppl 4):S50–S60.
4. Auvert B, Taljaard D, Lagarde E, et al.. Randomized, controlled intervention trial of male circumcision for reduction of HIV infection risk: the ANRS 1265 Trial. PLoS Med. 2005;2:e298.
5. Bailey RC, Moses S, Parker CB, et al.. Male circumcision for HIV prevention in young men in Kisumu, Kenya: a randomised controlled trial. Lancet. 2007;369:643–656.
6. Gray RH, Kigozi D, Serwadda D, et al.. Male circumcision for HIV prevention in men in Rakai, Uganda: a randomised trial. Lancet. 2007;369:657–666.
7. Ramjee G, Kapiga S, Weiss S, et al.. The value of site preparedness studies for future implementation of phase 2/IIb/III HIV prevention trials: experience from the HPTN 055 study. J Acquir Immune Defic Syndr. 2008;47:93–100.
8. Abdool Karim Q, Abdool Karim SS, Frohlich JA, et al.. Effectiveness and safety of tenofovir gel, an antiretroviral microbicide, for the prevention of HIV infection in women. Science. 2010;329:1168–1174.
9. Cohen MS, Chen YQ, McCauley M, et al.. Prevention of HIV-1 infection with early antiretroviral therapy. N Engl J Med. 2011;365:493–505.
10. Jaoko W, Nakwagala FN, Anzala O, et al.. Safety and immunogenicity of recombinant low-dosage HIV-1 A vaccine candidates vectored by plasmid pTHr DNA or modified vaccinia virus Ankara (MVA) in humans in East Africa. Vaccine. 2008;26:2788–2795.
11. Mugyenyi PN. HIV vaccines: the Uganda experience. Vaccine. 2002;20:1905–1908.
12. van de Wijgert J, Jones H. Challenges in microbicide trial design and implementation. Stud Fam Plann. 2006;37:123–129.
13. Temoshok LR. Behavioral research contributions to planning and conducting HIV vaccine efficacy studies. AIDS Res Hum Retroviruses. 1994;10(suppl 2):S277–S280.
14. Excler JL, Rida W, Priddy F, et al.. A strategy for accelerating the development of preventive AIDS vaccines. AIDS. 2007;21:2259–2263.
15. Excler JL, Rida W, Priddy F, et al.. AIDS vaccines and preexposure prophylaxis: is synergy possible? AIDS Res Hum Retroviruses. 2011;27:669–680.
16. Shattock RJ, Warren M, McCormack S, et al.. AIDS. Turning the tide against HIV. Science. 2011;333:42–43.
17. Fast PE, Excler JL, Warren M, et al.. Clinical site development and preparation for AIDS vaccine efficacy trials in developing countriesIn: Koff WC, Kahn P, Gust ID, eds. AIDS Vaccine Development. Norfolk, United Kingdom: Caister Academic Press; 2007:87–98.
18. Grant RM, Lama JR, Anderson PL, et al.. Preexposure chemoprophylaxis for HIV prevention in men who have sex with men. N Engl J Med. 2010;363:2587–2599.
19. Padian NS, McCoy SI, Balkus JE, et al.. Weighing the gold in the gold standard: challenges in HIV prevention research. AIDS. 2010;24:621–635.
20. Hayes R, Kapiga S, Padian N, et al.. HIV prevention research: taking stock and the way forward. AIDS. 2010;24(suppl 4):S81–S92.
21. Dunkle KL, Greenberg L, Lanterman A, et al.. Source of new infections in generalised HIV epidemics—authors' reply (invited correspondence). Lancet. 2008;372:1300–1301.
22. Kamali A, Ruzagira E, Karita E, et al.. Establishing large cohorts in Africa to prepare for HIV prevention clinical trials. Presented at: 14th International Workshop on HIV Observational Databases; March 25–27, 2010; Sitges, Spain.
23. Directorate: Epidemiology and Surveillance, National HIV and Syphilis Antenatal Sero-prevalence Survey in South Africa 2005. Pretoria, South Africa: South African Department of Health; 2006.
24. Middelkoop K, Myer L, Mark D, et al.. Adolescent and adult participation in an HIV vaccine trial preparedness cohort in South Africa. J Adolesc Health. 2008;43:8–14.
25. Gray GE, Allen M, Moodie Z, et al.. Safety and efficacy of the HVTN 503/Phambili study of a clade-B-based HIV-1 vaccine in South Africa: a double-blind, randomised, placebo-controlled test-of-concept phase 2b study. Lancet Infect Dis. 2011;11:507–515.
26. Kimani J, Kaul R, Nagelkerke NJ, et al.. Reduced rates of HIV acquisition during unprotected sex by Kenyan female sex workers predating population declines in HIV prevalence. AIDS. 2008;22:131–137.
27. Taegtmeyer M, Muhaari A, Davies A, et al.. Challenges in addressing counselling needs of MSM in highly stigmatized contexts: results of a qualitative study from Kenya. Presented at: 15th Annual Conference of the British HIV Association (BHIVA). Arena and Convention Centre; April 1-3, 2009; Liverpool, United Kingdom.
28. van Loggerenberg F, Mlisana K, Williamson C, et al.. Establishing a cohort at high risk of HIV infection in South Africa: challenges and experiences of the CAPRISA 002 acute infection study. PLoS One. 2008;3:e1954.
29. Kebba A, Imami N, Bugembe-Lule D, et al.. Recent HIV-1 infection in a high-risk Ugandan cohort: implications for Phase IIB test-of-concept HIV vaccine trials. Pharmacogenomics. 2007;8:409–414.
30. Baeten JM, Richardson BA, Martin HL, Jr, et al.. Trends in HIV-1 incidence in a cohort of prostitutes in Kenya: implications for HIV-1 vaccine efficacy trials. J Acquir Immune Defic Syndr. 2000;24:458–464.
31. Djomand G, Metch B, Zorrilla CD, et al.. The HVTN protocol 903 vaccine preparedness study: lessons learned in preparation for HIV vaccine efficacy trials. J Acquir Immune Defic Syndr. 2008;48:82–89.
32. Mwangome M, Okuku HS, Fegan G, et al.. MSM sex workers maintain a high HIV-1 incidence over time in Mombasa, Kenya. Presented at: 5th IAS Conference on HIV Pathogenesis, Treatment, and Prevention; July 19–22, 2009; Cape Town, South Africa.
33. Geibel S, van der Elst EM, King'ola N, et al.. ‘Are you on the market?’: a capture-recapture enumeration of men who sell sex to men in and around Mombasa, Kenya. AIDS. 2007;21:1349–1354.
34. van der Elst EM, Okuku HS, Nakamya P, et al.. Is audio computer-assisted self-interview (ACASI) useful in risk behaviour assessment of female and male sex workers, Mombasa, Kenya? PLoS One. 2009;4:e5340.
35. Sanders EJ, Graham SM, Okuku HS, et al.. HIV-1 infection in high risk men who have sex with men in Mombasa, Kenya. AIDS. 2007;21:2513–2520.
36. Smith A, Tapsoba P, Peshu N, et al.. Men who have sex with men and HIV/AIDS in sub-Saharan Africa. Lancet. 2009;374:416–422.
37. Beyrer C, Baral SD, Walker D, et al.. The Expanding Epidemics of HIV Type 1 Among Men Who Have Sex With Men in Low- and Middle-Income Countries: Diversity and Consistency. Epidemiol Rev. 2010;32:137–151.
38. van Griensven F, de Lind van Wijngaarden JW, Baral S, et al.. The global epidemic of HIV infection among men who have sex with men. Curr Opin HIV AIDS. 2009;4:300–307.
39. Geibel S, Luchters S, King'ola N, et al.. Factors associated with self-reported unprotected anal sex among male sex workers in Mombasa, Kenya. Sex Transm Dis. 2008;35:746–752.
40. McGowan I. Rectal microbicides: a new focus for HIV prevention. Sex Transm Infect. 2008;84:413–417.
41. Onyango-Ouma W, Birungi H, Geibel S. Understanding the HIV/STI prevention needs of men who have sex with men in Kenya. In: Horizons Research Summary. Washington, DC: Population Council; 2006.
42. UNAIDS. UNAIDS action framework: universal access for men who have sex with men and transgender people. Geneva, Switzerland: Joint United Nations Program on HIV/AIDS; 2009.
43. Beyrer C, Baral S, Kerrigan D, et al.. Expanding the Space: inclusion of most-at-risk populations in HIV prevention, treatment, and care services. J Acquir Immune Defic Syndr. 2011;57(suppl 2):S96–S99.
44. Benjamin Brown ZD, Scheibe A, Sanders E, eds. Men who have Sex With Men: An Introductory Guide for Health Care Workers in Africa Revised edition 2011. Cape Town, South Africa: Desmond Tutu HIV Foundation; 2011.
45. Braunstein SL, van de Wijgert J, Nash D. HIV incidence in sub-Saharan Africa: a review of available data with implications for surveillance and prevention planning. AIDS Rev. 2009;11:140–156.
46. Ahmed S, Lutalo T, Wawer M, et al.. HIV incidence and sexually transmitted disease prevalence associated with condom use: a population study in Rakai, Uganda. AIDS. 2001;15:2171–2179.
47. Riedner G, Hoffmann O, Rusizoka M, et al.. Decline in sexually transmitted infection prevalence and HIV incidence in female barworkers attending prevention and care services in Mbeya Region, Tanzania. AIDS. 2006;20:609–615.
48. Sateren WB, Foglia G, Renzullo PO, et al.. Epidemiology of HIV-1 infection in agricultural plantation residents in Kericho, Kenya: preparation for vaccine feasibility studies. J Acquir Immune Defic Syndr. 2006;43:102–106.
49. Bakari M, Lyamuya E, Mugusi F, et al.. The prevalence and incidence of HIV-1 infection and syphilis in a cohort of police officers in Dar es Salaam, Tanzania: a potential population for HIV vaccine trials. AIDS. 2000;14:313–320.

Africa; HIV risk; HIV incidence; retention; MSM; sexual behavior

© 2012 Lippincott Williams & Wilkins, Inc.