Introduction

Adolescent girls and young women (AGYW) aged 15–24 are at disproportionately high risk of acquiring HIV in sub-Saharan Africa; they accounted for 25% of new adult HIV infections in 2015 [1]. Young women who sell sex (YWSS) face particularly high risks due to high numbers of sexual partners, inability to negotiate condom use, and limited access to health services[2, 3]. Survey and programme data collected from female sex workers across Zimbabwe since 2009 suggest that that YWSS exist in sizeable numbers, and around 20% of female sex workers (FSW) report initiating sex work before age 20 [4–6]. Sex work is criminalised in Zimbabwe, although increasingly ignored by the police [7].

To reduce HIV among AGYW, the DREAMS initiative (Determined Resilient Empowered AIDS-free Mentored and Safe), targets biological, behavioural, and socio- economic determinants of risk through a comprehensive package of HIV prevention interventions in 10 African countries. In Zimbabwe, DREAMS is being implemented across six districts, delivered through local implementing partners that provide access to condoms, contraception, and HIV testing, and offer educational opportunities, vocational training, and work experience to address AGYW’s socio-economic vulnerabilities. Furthermore, DREAMS is making oral pre-exposure prophylaxis (PrEP) available to YWSS aged 18–24 in four of its programme districts in Zimbabwe to ameliorate their heightened risk.

To assess the effectiveness of the DREAMS initiative over three years, two intervention (Bulawayo and Mutare) and four comparison (Chinhoyi, Karoi, Kwekwe and Zvishavane) sites were selected for inclusion in impact and process evaluation studies [8]. Prior to baseline data collection and introduction of the DREAMS package, we conducted a rapid needs assessment in each of the six sites to deepen understanding of the local context. For both intervention and comparison sites, knowledge of where and how YWSS sell sex, their relationships with each other and with adult FSW, whether they are mobile between the sites or remain “clustered”, and how they themselves perceive their involvement in exchanging sex for money or other material goods was a prerequisite for the appropriate intervention and evaluation design.

In this paper, we describe the “social mapping” approach used during the rapid needs assessment that combined identification of geographical locations where YWSS sell sex with additional data on the social context in which sex is exchanged. The aim of the study was to “paint a picture” of the YWSS physical and social environment in each study site to prepare for the DREAMS intervention and evaluation. Specific objectives of social mapping were to: 1) understand the organization and structure of commercial sex provided by YWSS and effects on their vulnerability; 2) explore YWSS social networks and levels of support; and 3) determine whether YWSS would be willing to participate in research activities including baseline and follow-up surveys and, in the intervention sites, their interest in accessing activities planned as part of DREAMS. By describing how this structured, rapid appraisal identified similarities and differences between sites, we highlight key lessons that could be useful for other programmes attempting to reach this marginalised and often hidden key population.

Methods

We conducted social mapping in all sites selected for research to evaluate delivery and effectiveness of the DREAMS programme, including two DREAMS implementation sites (the cities of Bulawayo and Mutare) and four smaller towns designated as comparison sites (Chinhoyi, Karoi, Kwekwe and Zvishavane). All six sites provide targeted clinical services, peer education, and community mobilisation meetings for FSW as part of the Sisters with a Voice national sex worker HIV prevention and treatment programme. The Sisters programme is delivered by the Centre for Sexual Health, HIV and AIDS Research (CeSHHAR), which is also one of the organisations responsible for recruiting and referring YWSS into DREAMS activities, where available. In collaboration with the London School of Hygiene and Tropical Medicine, CeSHHAR will conduct the process and impact evaluation of DREAMS Zimbabwe.

We collected social mapping data across each site, comprising the city or town’s central business district (CBD) and all surrounding residential and industrial areas mentioned as sites where YWSS sell sex. We defined YWSS as women aged between 18–24 who exchanged sex for money or material goods, including those who self-identify as FSW as well as those who do not consider themselves as sex workers but report sex with men that would not occur in the absence of receiving financial or other resources. DREAMS targets the 10–24 age group, but only YWSS aged 18 and above are eligible for PrEP. The Zimbabwe Medical Research Council allows young people aged eighteen and above to provide consent to take part in research. Ethical approval for the DREAMS evaluation research was obtained from the Medical Research Council of Zimbabwe (MRCZ/2085) and the London School of Hygiene and Tropical Medicine (11835).

The mapping process started with informal discussions with locally deployed peer educations from the Sisters programme to gain an initial understanding of the geographical spread of YWSS, particularly where YWSS could be found at different times. Once a basic geographical picture was built up, a team of 3–4 female fieldworkers visited each site mentioned. Fieldworkers were CeSHHAR research staff with extensive experience of both quantitative and qualitative data collection with sex workers.

Venues included bars, hotels, guesthouses and lodges, nightclubs, markets, streets, bus/truck stops and taxi ranks. The fieldwork team visited each suggested location on the three “busiest days” of the week, Thursday, Friday and Saturday over two consecutive weekends. During site visits, we used three rapid assessment methods: direct observation, group discussions, and informal interviews. Table 1 summarises information collected by each method.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Data collection methods.

https://doi.org/10.1371/journal.pone.0194301.t001

Fieldworkers were trained to naturally engage with the “scene” at each site, and to approach people in a friendly manner, providing their affiliation with the Sisters programme, which was familiar to many FSW and thus helped reduce anxiety. Fieldworkers explained the purpose of the social mapping and briefly chatted to all individuals encountered during the observation. Conversations were conducted in the local language (Shone or Ndebele) and notes were taken throughout the observation. No identifying information was recorded. Fieldworkers were encouraged to conduct “natural” conversations without topic guides. In addition to YWSS, we talked to key informants from the community including older FSW, college and university students, owners or employees of bars and other large entertainment venues (including nightclubs), hotels and guesthouses, street vendors (particularly near bars), taxi drivers, and health service staff/NGO workers who provide services to sex workers. All informants were asked to suggest additional locations we could visit to observe YWSS.

During these observations we identified YWSS to take part in group discussions and individual interviews. These discussions included both young women who openly sell sex and those who do not self-identify as sex workers but explicitly exchange sex for money or material goods. One discussion was held in each of Mutare (n = 17), Zvishavane (n = 7), Kwekwe (n = 6), Chinhoyi (n = 10) and Karoi (n = 7). Two groups were organised in Bulawayo (n = 25+15). Individual interviews were conducted with young women identified by peer educators and other community informants, and were selected for interview because of their openness during group discussions, irrespective of whether they had previously attended the Sisters clinic. They were recruited from different venues, reflecting different means for locating clients or boyfriends. All participants for group discussions and interviews provided verbal consent. As before, only written notes were taken for analysis and no identifying information recorded.

In keeping with the rapid nature of the assessment, we conducted basic content analysis of notes from observations, group discussions and interviews over several days following social mapping. For each site, we developed a detailed matrix depicting characteristics of the different venues and YWSS working there, and summarizing the number and type of respondents from whom data were collected, to ensure reliability across sites. A de-identified example is provided as Table 2, to avoid inadvertent disclosure of specific venues where YWSS sell sex which could then be targeted by unwanted publicity or police action. These tables were used to help plan for appropriate ways of reaching YWSS in intervention sites for referral into DREAMS, and to check for basic comparability across intervention and comparison sites.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Sample YWSS venue observation matrix.

https://doi.org/10.1371/journal.pone.0194301.t002

Results

YWSS sell sex at diverse locations and events, some of which are at the boundary of sex work and more informal transactions/ exchanges. We found a continuum in how YWSS perceived their involvement in sex work, from those who identified as sex workers to those who did not consider their sexual exchanges to be income-generating activities so much as social events with financial and material benefits. We present our findings according to whether they were common to all study sites or specific to one or more, as this has implications for comparability across sites and over time during the DREAMS evaluation study. As we move forward toward evaluating DREAMS uptake and effects in each site, and compare these to non-DREAMS locations, we will need to consider how programme components interact with the local YWSS context.

We highlight findings on geographical location, YWSS characteristics, the organisation of sex work, social relationships, and uptake of and interest in services as dimensions of our “social maps”. Fig 1 provides a visual depiction of the kind of geographical and socio-demographic data compiled for each location.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. An illustration showing the geographic and sociodemographic data compiled at each site during social mapping.

https://doi.org/10.1371/journal.pone.0194301.g001

Discussion

We used “social mapping” as a structured rapid appraisal approach to characterise the YWSS context in six sites over a short period of time, gaining understanding of both geographical and social patterns of how younger women sell sex. We conducted the mapping to collect practical information to inform strategies for ensuring the DREAMS programme reaches YWSS among the broader target population of high risk young women. Our results are also salient for the design of process and impact evaluations, which will need to accommodate differences in how YWSS sell sex, interact, and seek services.

We found many similarities across sites. Most YWSS live outside their city or town’s CBD and travel into town on weekend evenings once their local bars close. They do so to both preserve anonymity and seek out higher paying clients. Those who remain in the high density suburbs charge less money and see themselves as professional FSW. Due to hostility from older sex workers, YWSS avoid bars and other locations with a concentration of FSW, instead relying on attracting clients outside them, taking advantage of looking younger and “new”. This approach, coupled with their tendency to charge less money or accept food or other material goods in lieu of cash, contributed to ongoing conflicts with older FSW. YWSS also fear being identified and hide their work from parents and other community members, making them reluctant to associate closely with one another. Fear of disclosure appears to make them less willing to seek sexual health care, even at targeted services for sex workers such as Sisters clinics.

We also identified several important contextual differences across our six study sites. Most notably, the four university towns had unique opportunities for YWSS to transact sex, and university students were the least likely to self-identify as sex workers or perceive their engagement in transactional sex to be high risk. Students did not self-identify as sex workers, and did not congregate in venues typically associated with commercial sex. Nonetheless, their client base extended beyond the student population, potentially overlapping with clients of other sex workers. Perhaps at the other end of the YWSS continuum, migrant young women from rural areas perceived themselves to be school “drop outs” and “runaways” and were more likely to see themselves as professionalised sex workers, seeking out groups of men known to be clients such as illegal gold panners or truck drivers.

Some of our findings reflect existing evidence on YWSS from Zimbabwe and other contexts. For example, YWSS often move from less formal transactional sexual exchange into more highly structured and “professional” sex work over time [9, 10]. They engage in sex work for multiple reasons; some require significant livelihood support, while others seek luxury gifts or disposable income [11–14]. Younger sex workers tend to be fearful of engaging with services out of fear of their involvement becoming known to parents or other community members [15, 16]. Furthermore, hostility between older and younger sex workers exacerbates competition and makes it more difficult for them to collaborate for improving work conditions and reducing vulnerability. For example, in Uganda, younger sex workers also avoided contact with older FSW by working outside bars [17]. Participation of school and university students in transactional sex is widespread and many studies have similarly found students to be unlikely to consider their exchange of sex for financial or material benefits to be sex work per se [12, 18, 19].

Our findings also provide new insights on YWSS and their vulnerability to HIV, particularly for the Zimbabwean context. Although CeSHHAR has worked with FSW across the country since 2009, this study is the first time we came across formally organised parties at which YWSS are recruited by same-age or older males specifically for arranging sexual exchange. Both the Chillers parties and the university functions integrate social events for young people with organising transactional relationships with older males. Both events combine on-site, short term sexual encounters with opportunities for YWSS to recruit longer term “blessers” or regular clients. In general, the way in which YWSS sell sex is poorly understood and often excluded from the literature on FSW due to the considerable ethical considerations in including YWSS under age 18 in research [20, 21]. Increasingly, however, there have been calls to move beyond simplistic portrayals of all underage YWSS as “trafficking victims” in order to gather data that are required for effective methods of reaching this vulnerable population, including offering services that are likely to be acceptable and taken up by YWSS, rather than driving them further underground and into higher risk situations [10, 21, 22].

There are thus several implications of our “social mapping” exercise for the DREAMS programme and its evaluation. A comprehensive understanding of the local context in which YWSS work, live, and seek opportunities to improve their lives will help guide suitable messages and interventions. First, even where there are common trends in how YWSS live and work, they remain largely outside existing FSW social networks and do not engage with targeted services. Linking them to DREAMS’ health and social interventions will require intensive measures and cannot rely on leveraging adult FSW peer educators and their familiarity with the community. Second, there are clearly different typologies of YWSS, for example “street based” women who nonetheless cluster in known sex work “hot spots”, female university students who transact sex during term times at social events, and migrant rural AGYW who target concentrations of men with disposable income who are themselves difficult to identify, such as illicit gold panners. These groups of YWSS do not associate with each other, nor would them all self-identify as YWSS to the same degree. Approaches used to engage young women aged 18–24 who consider themselves professional sex workers will need to differ from those who consider their exchange of sex for money or goods to be part of a temporary livelihood strategy. Care needs to be taken in considering which components of the DREAMS package are likely to appeal to each group, and how to ensure referrals are taken up. For example, self-described school drop-outs may find educational subsidies or vocational training the most appealing entry point into DREAMS, while YWSS engaged in selling sex on the street may want youth-friendly clinical sexual and reproductive health services in the first instance. Students may be split between those interested in educational support, job opportunities, or “edu-tainment” type HIV awareness raising campaigns. Following initial, positive contact with DREAMS may subsequently lead to higher uptake of a broader constellation of services.

Finally, the evaluation of DREAMS requires that the baseline context of vulnerability be clearly identified in both intervention and comparison sites, to allow for comparison across time and place. Due to the way DREAMS was designed for Zimbabwe, it was difficult to select similar sites for intervention and comparison. For this reason, the two urban areas where the intervention is being implemented have been paired with 4 smaller towns for comparison. While the numbers of YWSS may be broadly similar, the nature of how they sell sex has some significant differences that need to be taken into account.

Conclusions

Successful intervention and research design rely on in-depth understanding of local contexts and risk environments. Our social mapping of YWSS across six sites in Zimbabwe, provides one example of applying a structured yet informal rapid appraisal to gather information useful to subsequent recruitment of a hard-to-reach population. Our YWSS matrix and (sample) visual map demonstrate the kinds of outputs generated and may provide a starting point for research into YWSS in other contexts.

Acknowledgments

We would like to thank the DREAMS Impact Evaluation research team and Sisters with a Voice staff for their assistance in contacting the research participants. We thank the participants for agreeing to take part in this study.