SEXUALLY TRANSMITTED INFECTIONS (STIs) constitute a significant public health concern in Madagascar. Surveys conducted in the mid-1990s highlighted high syphilis and low but rising HIV prevalence levels.1,2 More recently, high prevalence of gonorrhea, Chlamydial infection, trichomoniasis, or syphilis was measured among asymptomatic female sex workers (74%–78%)3 and women seeking care for vaginal discharge (40%).4 STI-related morbidity warrants a public health response. Given the role of STIs in HIV transmission, particularly in a nascent HIV epidemic,5,6 STI control is also urgently needed to help prevent a widespread HIV epidemic in Madagascar.5,7,8
Persistent STI indicates current control strategies have been inadequate and novel approaches are needed. The Priorities for Local AIDS Control Efforts (PLACE) Study is an STI/HIV intervention-planning tool based on epidemiologic models indicating that new and multiple sexual partnerships are important STI/HIV transmission determinants.9,10 PLACE identifies geographic areas and social venues within these areas where people meet new and multiple partners and measures the unmet need and potential for condom promotion and STI/HIV prevention education at social venues. A venue-based approach can offer programmatic advantages over approaches based on risk group status that can be stigmatizing and that may be inadequate in generalized epidemics, such as that observed in Madagascar.
A research team including representatives from the Madagascar National AIDS Control Program; the Madagascar National Institute for Public and Community Health; the United States Aid for International Development; the University of North Carolina, Madagascar; and the MEASURE Evaluation Project at the University of North Carolina, Chapel Hill implemented the Madagascar PLACE Study in 7 cities. We aimed to identify venues where people meet new sexual partners, assess the potential for STI/HIV intervention hosted at venues, and measure the prevalence of high-risk sexual behaviors and self-reported STIs among individuals socializing at venues.
Materials and Methods
The research team identified 7 priority towns where population mobility was high and the potential for sexual partnership and STI/HIV risk was thought to be likely: agricultural and industrial center Antsirabe (population: 162,000 inhabitants); urban port towns Mahajanga, Morondava, and Mananjary (population: 166,000, 70,000, and 70,000 inhabitants, respectively); coastal tourist town and mining area, Fort Dauphin (population: 50,000 inhabitants); semirural town, Tsiroanomandidy, which hosts a large zebu (cattle) market weekly (population: 50,000 inhabitants); and sapphire mining settlement Ilakaka, reputed for its sex work industry (population:15,000 inhabitants).
Field work was conducted from May through December 2003. The PLACE Study methodology has been described in detail elsewhere.11 Briefly, in the first phase, community informants in each study town were interviewed to attempt to identify all social venues in the town where people meet new sexual partners. In the second phase, each identified venue was visited to verify the venue address and administer a 10 minute structured face-to-face questionnaire to a venue representative, such as a venue manager or owner, to assess the history of and the potential for STI/HIV intervention based in the social venues. In the final phase, in each town, 20 venues were chosen randomly with a probability of selection that was proportional to the size of the on-site socializing population, measured during interviews with venue representatives. An additional 5 priority venues per town were purposefully selected by the field coordinator. Fifteen minute structured face-to-face sexual behavior surveys were administered to a sample of venue patrons recruited from the 25 venues. Sixteen men and 8 women were recruited at each venue, as interviews with venue representatives indicated a venue sex ratio of 2 men to 1 woman. A protocol was followed that distributed interviewers systematically throughout the venue to minimize interviewer discretion in selecting respondents by convenience. Respondents were led to a private area to ensure confidentiality during the interview.
Before each interview, interviewers explained the purpose and scope of the study and obtained a verbal informed consent to conduct a confidential and anonymous interview. The interviewer provided a small snack bag to each respondent to eat during the interview. All interviews were conducted in Malagasy.
Interviewers were selected based on skill, experience, and fluency in Malagasy. Each interviewer received a 1-week training on the PLACE Study protocol. Study instruments were translated from English into Malagasy and French.
The ethical review boards of Madagascar Medical Research and the University of North Carolina at Chapel Hill approved the research.
Venue STI/HIV Intervention.
During the interviews with venue representatives, STI/HIV intervention indicators were measured, including prior on-site STI/HIV prevention intervention, prior on-site condom availability, and venue representative willingness to host STI/HIV prevention education and sell condoms.
Venue Patron Characteristics.
Measured characteristics of venue patrons included age, employment, education, and exposure to STI/HIV education. Sexual behavior indicators included how often the patron visited the venue, new sexual partnerships, transactional sex, and condom use. Transactional sex was defined as having given or received money, goods, or services for sex in the past 4 weeks. Typically in Madagascar, women sell and men buy sex, hence this variable indicated whether a woman was a sex worker or a man was the client of a sex worker. Patrons also self-reported STI symptoms in the past 4 weeks.
Analyses were performed in Stata, version 8.0 (Stata Corp., College Station, TX). The number of venues was enumerated and the distributions of major venue types and on-site STI/HIV intervention variables were calculated separately by study town.
Patron socio-demographic characteristics, sexual behaviors, and self-reported STIs were calculated separately by gender and compared using Pearson chi-square test statistics.
To compare STI risk among individuals involved in sex trade versus other high-risk partnerships (without sex trade), we measured levels of symptoms suggestive of STI among those reporting that, in the past 4 weeks, they engaged in: transactional sex and at least 1 new partnership; transactional sex, but no new partnerships; at least 1 new partnership, but no transactional sex; and no new partnerships and no transactional sex. We calculated gender-specific unadjusted and adjusted prevalence ratios (PRs) and 95% confidence intervals (CIs) for the associations between risk behavior and STI symptoms using a generalized linear model specifying a log link, a Poisson distribution,12,13 and a robust variance estimator.14 Multivariable models included study town, age, and employment.
We calculated unadjusted and adjusted PRs and 95% CIs for the associations between exposure to STI/HIV prevention and 2 indicators of STI risk: condom use with a recent new sexual partner and self-reported STI symptoms in the past 4 weeks. Results were not gender-stratified as preliminary analyses indicated few differences by gender. Multivariable models adjusted for study town, gender, age, employment, and transactional sex.
Number and Types of Venues: Community Informant Interviews
The number of community informants interviewed ranged from 147 in Morondava to 201 in Antsirabe (response rates: 98%–100%, depending on the town). Informants were diverse and included merchants, taxi drivers, restaurant/hotel workers, and people in the street. Numerous venues where people meet new sexual partners were identified (range: 67–211 venues) (Table 1). Informants identified the greatest number of venues in the largest port cities, Mahajanga, Morondava, and Mananjary (211, 171, and 106 venues, respectively). The most common venues were open-aired public spaces such as streets, parks, beaches, or fields (range: 30% of venues in Fort Dauphin and Ilakaka to 49% of venues in Antsirabe) followed by hotel/bar establishments (range: 13% of venues in Tsiroanomandidy to 49% of venues in Fort Dauphin). Additional venues included private homes, many of which functioned as informal brothels; informal shacks serving meals, called gargottes; schools; video clubs; transport centers, such as taxi stands and bus stations; shops and markets; and churches. Setting-specific venues included mines in Ilakaka and Fort Dauphin, the zebu market in Tsiroanomandidy, and casinos in Antsirabe.
STI/HIV Prevention Activities at Venues: Venue Representative Interviews
Interviewers identified a venue representative to approach for an interview at the majority of venues in each study town (range: 91%–100%, depending on the town). Interviews with 644 representatives (response rates: 92%–100%) indicated that on-site STI/HIV prevention activities had previously been hosted in approximately one quarter of venues in Mahajanga (24%) and Antsirabe (25%), 33% of venues in Mananjary, 39% of venues in Morondava, and between 43% and 45% of venues in Tsiroanomandidy, Ilakaka, and Fort Dauphin (Table 2). Condoms were available on the day of the interview at less than one third of venues in Tsiroanomandidy (27%), Mananjary (27%), Morondava (30%), and Mahajanga (32%), 41% of venues in Antsirabe, and 54% of venues in Fort Dauphin.
The majority of venue representatives were willing to host prevention activities at their venues, or would consider the option (range: 73%–90%). Smaller proportions indicated they would be willing to or would consider selling condoms, with commitment lower in Tsiroanomandidy (48%), Mananjary (50%), and Morondava (55%) than in Mahajanga (68%), Antsirabe (77%), Fort Dauphin (79%), and Ilakaka (83%).
Sexual Behavior Survey Among Venue Patrons
Socio-Demographic Characteristics, STI/HIV Education, and PLACE Venue Visitation.
Of 2021 men and 977 women aged 18 years or older who were approached for an interview while socializing at the venue, 2013 (99.6%) men and 969 (99.2%) women agreed to the interview.
Women interviewed at venues were younger than their male counterparts (median of 23 versus 26 years, respectively) (Table 3). Women were less likely to report any employment, whether full- or part-time (49%), than men (73%) (P < 0.001). One fifth of men and women were currently students.
Recent exposure to STI/HIV prevention intervention was comparable among men and women (Table 3). Only 28% of men and women had participated in an STI/HIV education in the past 3 months. However, the majority had heard a radio program (men: 90%, women: 88%), seen a public health advertisement (men: 85%, women: 83%), or watched a TV program (men: 68%, women: 63%) about STI/HIV in the past 3 months.
Approximately half of respondents reported daily visits to the PLACE social venues (50% men, 51% women).
Sexual Behavior and STI Symptoms.
In the past 4 weeks, 78% of men and 74% of women reported having either at least 1 new sexual partner or engaging in transactional sex (Table 3). The greatest percentage of respondents reported both transactional sex and multiple new sexual partnerships (men: 45%, women: 47%). Small percentages of respondents reported transactional sex but no new sexual partnerships (men: 16%, women: 14%) or new sexual partnerships but no transactional sex (men: 14%, women: 9%). The distributions of high-risk partnerships differed somewhat by gender (P = 0.003). More striking, however, was that women were more likely than men to report very high numbers of recent new sexual partners; nearly 7% of women reported 20 or more new partners in the past 4 weeks compared with less than 1% of men (P < 0.001).
Approximately half of respondents reported ever having used a condom (men: 49%, women: 52%). Among persons with a new partner in the past 12 months, 28% of men and 41% of women used a condom the last time they had sex with a new partner (P < 0.001).
Approximately 19% of men and 18% of women reported symptoms suggestive of an STI in the past 4 weeks, including pain on urination (men), discharge from the penis (men), unusual vaginal discharge (women), lower abdominal pain (women), and/or genital ulcers (men and women). Of those reporting STI symptoms, the majority reported having consulted a doctor in response to the symptoms (men: 71%, women: 67%). Whether the doctor was a traditional healer or a medical doctor was not asked in the interview.
Some differences in levels of sexual risk behavior by study town were observed (not presented in table). Although 46% of men and women overall reported both new sexual partnerships and engaging in transactional sex in the past 4 weeks, these high-risk sexual partnerships were disproportionately high in Mananjary (61%), Ilakaka (56%), and Fort Dauphin (49%). Condom use with a recent new partner was low for men and women overall (32%), though levels were particularly low in Ilakaka (16%), Tsiroanomandidy (25%), and Fort Dauphin (27%).
Associations Between Risk Behaviors and STI Symptoms.
Compared with those who reported neither transactional sex nor a new sexual partnership in the past 4 weeks, the prevalence of self-reported STI symptoms in the past 4 weeks was greater among those reporting both transactional sex and at least 1 new partnership (men: unadjusted PR: 4.02, 95% CI: 2.84–5.70; women: 3.77, 95% CI: 2.32–6.12), transactional sex but no new sexual partnerships (men: unadjusted PR: 1.67, 95% CI: 1.07–2.61; women: 2.77, 95% CI: 1.55–4.94), and at least 1 new sexual partnership but no transactional sex (men: unadjusted PR: 2.26, 95% CI: 1.47–3.46; women: 2.33, 95% CI: 1.20–4.52) (Table 4). After adjusting for study town, age, and employment status, these estimates weakened but the associations generally persisted.
Associations Between Prior Exposure to STI/HIV Interventions and Current Risk of STI.
Individuals who had been exposed to STI/HIV prevention education materials in the past 3 months were more likely to report recent condom use with a new sexual partner and less likely to report STI symptoms in the past 3 months (Table 5). For example, those who had heard a radio program about STI/HIV in the past 3 months, compared with those who were not exposed to radio messages, were twice as likely to report recent condom use (unadjusted PR: 1.98, 95% CI: 1.47–2.66) and less likely to report recent STI symptoms (unadjusted PR: 0.70, 95% CI: 0.57–0.87). After adjusting for study town, gender, age, and employment status, and involvement in transactional sex, the associations weakened somewhat but remained.
Approximately three-quarters of this venue-based sample reported having at least 1 new sexual partnership or engaging in transactional sex in the past 4 weeks. An important condom use gap was observed given these high partnership levels; approximately one half of the respondents reported ever having used a condom and a minority used a condom the last time they had sex with a new sexual partner. Self-reported symptoms suggestive of STIs were common. These results suggest that venue patrons represent a population at increased STI risk in Madagascar.
PLACE venues could play an important role in STI/HIV prevention in Madagascar. A minority of identified venues had previously hosted STI/HIV prevention or had on-site condom availability in the year before the interview. However, between 70% and 90% were open to hosting prevention activities and the majority were open to selling condoms. Appropriate interventions should be developed for these venues where otherwise difficult-to-reach populations vulnerable to infection can be accessed.
Individuals who socialized at venues included female sex workers and their male clients. Some women reported very high numbers of recent new partnerships, highlighting the particular vulnerability to STI among this population. Men and women involved in transactional sex indeed reported disproportionate levels of STI symptoms. However, respondents who reported recent new partnerships but no transactional sex were also more likely to report STI symptoms than those reporting no new partnerships and no transactional sex. This finding indicates that STI/HIV prevention efforts focused on sex worker populations may fail to reach some individuals at risk of STI/HIV. Interventions implemented in PLACE venues, however, would reach a spectrum of high-risk populations.
Our results indicate that, among venue patrons, prior exposure to STI/HIV prevention was associated with increased condom use and lower STI symptom levels. The causal effect of prevention interventions on venue patron behavior could not be evaluated based on these cross-sectional data. Nonetheless, the findings pointed to the potential positive effects of prior STI/HIV prevention interventions. We believe expansion of STI/HIV prevention interventions, including into the venue environment, is warranted and feasible. Radio programs that play music and include messages about STI/HIV prevention could be broadcast at bars and clubs, whereas educational posters could be hung in nearly any venue.
Low levels of condom use among the sample point to the need for condom promotion. Venue-based condom promotion has obvious potential, given experiences in Thailand,15 Nicaragua,16 and the southern United States.17 A condom promotion trial among sex workers in Madagascar indicated that an intervention of intensive clinic-based counseling plus peer education was associated with slightly lower STI incidence at 6 months versus peer education alone. However, 6 months after receiving presumptive treatment, the combined prevalence of Chlamydial infection, gonorrhea, or trichomoniasis in the intervention group was 32%, indicating that this condom promotion strategy alone was an inadequate STI prevention intervention.18 Clinic-based efforts coupled with condom promotion implemented in venues where people meet partners—at a time when condoms will soon be needed—may improve efforts to reduce STIs among sex workers, their clients, and other high-risk populations.
This PLACE Study attempted to identify the universe of venues where people meet sexual partners in each town, hence yielding a comprehensive list of places where high-risk groups can be reached. Further, participation levels were excellent, as has been observed in previous studies conducted in Madagascar by members of our team,3,19 suggesting nonresponse bias was minimized.
However, a number of study limitations should be noted. Venue information may be incomplete if community informants failed to name important venues or venue representatives were not fully knowledgeable about their venues and provided inaccurate information. In addition, the sexual behavior findings measured among venue patrons may not be fully representative of the entire population socializing at venues, because of a data entry error that resulted in the inability to separate the 20 randomly chosen venues from the 5 purposefully chosen venues. Finally, social desirability and recall biases may have influenced self-reported sexual behaviors,20,21 whereas self-reported STI served only as a crude indicator of the true burden of infection among venue patrons.
Although high-risk behaviors and symptoms suggestive of STIs were common among the venue population, venue managers were willing to host interventions. The study findings were discussed in workshops involving stakeholders and policy makers with the goal of outlining effective local strategies for STI/HIV prevention. Population Services International, an organization that has worked to increase condom access in Madagascar, has implemented condom promotion in some PLACE venues. STI/HIV interventions in venues identified during the PLACE study, locations where high-risk populations socialize, should be expanded to reach Malagasy men and women who are clearly in need of prevention interventions.
1. Behets FM, Andriamahenina R, Andriamiadana J, et al. High syphilis and low but rising HIV seroprevalence rates in Madagascar. Lancet 1996; 347:831.
2. Xueref S, Holianjavony J, Daniel R, et al. The absence of HIV seropositivity contrasts with a high prevalence of markers of sexually transmitted infections among registered female sex workers in Toliary, Madagascar Trop Med Int Health 2003; 8:60–66.
3. Behets FM, Rasolofomanana JR, Van Damme K, et al. Evidence-based treatment guidelines for sexually transmitted infections developed with and for female sex workers. Trop Med Int Health 2003; 8:251–258.
4. Behets F, Andriamiadana J, Rasamilalao D, et al. Sexually transmitted infections and associated socio-demographic and behavioural factors in women seeking primary care suggest Madagascar's vulnerability to rapid HIV spread. Trop Med Int Health 2001; 6:202–211.
5. Grosskurth H, Mosha F, Todd J, et al. Impact of improved treatment of sexually transmitted diseases on HIV infection in rural Tanzania: Randomised controlled trial. Lancet 1995; 346:530–536.
6. Korenromp EL, White RG, Orroth KK, et al. Determinants of the impact of sexually transmitted infection treatment on prevention of HIV infection: A synthesis of evidence from the Mwanza, Rakai, and Masaka intervention trials. J Infect Dis 2005; 191(suppl 1):S168–S178.
7. Cohen MS. HIV and sexually transmitted diseases: Lethal synergy. Top HIV Med 2004; 12:104–107.
8. Fleming DT, Wasserheit JN. From epidemiological synergy to public health policy and practice: The contribution of other sexually transmitted diseases to sexual transmission of HIV infection. Sex Transm Infect 1999; 75:3–17.
9. Anderson R. Transmission dynamics of sexually transmitted infections. In: Holmes KK, Sparling PF, Mardh PA, et al., eds. Sexually Transmitted Disease, 3rd ed. New York: McGraw-Hill, Health Professions Division, 1999.
10. Anderson RM, May RM. Epidemiological parameters of HIV transmission. Nature 1988; 333:514–519.
11. Weir SS, Pailman C, Mahlalela X, et al. People to places: Focusing AIDS prevention efforts where it matters most. AIDS 2003; 17:895–903.
12. McNutt LA, Wu C, Xue X, et al. Estimating the relative risk in cohort studies and clinical trials of common outcomes. Am J Epidemiol 2003; 157:940–943.
13. Zou G. A modified poisson regression approach to prospective studies with binary data. Am J Epidemiol 2004; 159:702–706.
14. Zocchetti C, Consonni D, Bertazzi PA. Estimation of prevalence rate ratios from cross-sectional data. Int J Epidemiol 1995; 24:1064–1065.
15. Rojanapithayakorn W, Hanenberg R. The 100% condom program in Thailand. AIDS 1996; 10:1–7.
16. Egger M, Pauw J, Lopatatzidis A, et al. Promotion of condom use in a high-risk setting in Nicaragua: A randomised controlled trial. Lancet 2000; 355:2101–2105.
17. Kelly JA, St Lawrence JS, Diaz YE, et al. HIV risk behavior reduction following intervention with key opinion leaders of population: An experimental analysis. Am J Public Health 1991; 81:168–171.
18. Feldblum PJ, Hatzell T, Van Damme K, et al. Results of a randomised trial of male condom promotion among Madagascar sex workers. Sex Transm Infect 2005; 81:166–173.
19. Behets F, Turner AN, Van Damme K, et al. Acceptability and feasibility of continuous diaphragm use among sex workers in Madagascar. Sex Transm Infect 2005; 81:472–476.
20. Catania JA, Gibson DR, Chitwood DD, et al. Methodological problems in AIDS behavioral research: Influences on measurement error and participation bias in studies of sexual behavior. Psychol Bull 1990; 108:339–362.
21. Geary CW, Tchupo JP, Johnson L, et al. Respondent perspectives on self-report measures of condom use. AIDS Educ Prev 2003; 15:499–515.