Procedure.

All PGDs were conducted in venues provided by CSOs. The discussion sessions lasted for about 90–120 minutes, were moderated by trained researchers and were audio-recorded. Participants were familiar with these settings as CSOs provided HIV-related services including HIV testing and family planning services. Before the beginning of fieldwork, graduate research assistants (RAs) responsible to facilitate the PGDs were trained on the protocol, discussion guide in both English (S1 Text) and Kiswahili (S2 Text) as well as principles of research involving human subjects and informed consent. All RAs were Swahili speakers and had experience conducting qualitative studies among key and vulnerable populations including FSWs. RAs built rapport with study participants and assumed neutrality by distancing themselves from program implementers (i.e., they declared to the participants that they were researchers and not part of the Sauti Project implementers).

A PGD guide with open-ended questions was used to facilitate the sessions. The discussion guide covered personal experiences with HIV testing services before and during access to Sauti Project services. Themes in the guide included factors that influenced the use of HIV testing and reasons for accepting or avoiding linkage to HIV care facilities.

We conducted one PGD in Dar es Salaam to determine the suitability of the questions asked and practical issues with conducting discussion sessions with FSWs. After it was clear that no significant changes were needed in the guide or procedure, thus the data collected was included in the analysis. During data collection, debriefing sessions were held between research assistants and senior researchers after every PGD session to discuss the depth of the data collected and emerging issues. We obtained written informed consent from all participants before beginning the group sessions. To protect the identity of the participants, we asked them to use pseudonyms during consenting. Numbers were assigned to every participant for reference during discussion.

The role of peer-educators.

Participants identified peer educators as the primary source of information on available HIV care services through the Sauti Project and facility-based HIV services. According to participants, availability of peer educators encouraged FSWs to go for HIV testing, and it influenced their decision to accept a referral to an HIV care clinic for further services. FSWs reported that when someone is sick, she usually seeks advice from peers about the illness and the peers recommend where to go for medical consultations. For example, the types of services offered, where to get them, whom to contact for help at times of need, and benefits of early initiation of ART.

Peer-escorted referral to treatment services was an essential influence on FSWs’ decision to accept their first HIV clinical visit to the health facilities. Participants indicated that escorted referral helped clients to receive the service quickly, and made them feel cared. FSWs pointed out that escorted referral increased their confidence and helped to reduce fears, especially during the first HIV clinic visits. For example, in one PGD a participant described that:

The good thing is that the home-based counsellors and peer educators knew the service providers, so before you go to the facility, they have already talked to the providers and you get the service immediately. However, when you are alone, you are challenged. First, you ask yourself how should I ask about the HIV service …[laughs] … there are many nurses and doctors there…you may end-up leaving without seeing them [laughs] … if you go alone you wait for a long time because no one knows you and you are afraid to ask

Facilitation of transport during the first HIV clinical visit.

Participants applauded about transport facilitation to attend the first HIV clinical visit for FSW to minimize some inconveniences e.g. financial constraints to seek services. Participants reported that sometimes FSWs who may have tested HIV-positive or have doubt about their HIV status, fail to present themselves to an HIV care clinic for further HIV services because they cannot afford the transport costs. To them, the facilitation of transport to the clinic was an important motivation for linkage to care.

Presence of project-linked FSW-friendly health care providers at the health facilities.

Participants pointed out that the friendly provider-client relationship had a significant impact on linkage to care. Most participants reported that the presence of service providers trained by Sauti Project to provided FSW-friendly services and motivated FSWs to attend HIV care services. According to some FSWs, health workers who received training from the programme were ‘less stigmatising’, ‘knowledgeable about FSWs’ health needs’, ‘approachable’, and ‘confidential’. The importance of the presence of friendly health care providers was emphasised in all PGDs. In one PGD a participant said:

For us (FSWs), it does not matter where you get the service (HIV care services) … what matters is … does the nurse or doctor understand me? How does she/he look at me? If I tell my stories about my clients and my situation [infection], is she ready to treat me nicely and will not tell other people that I have the disease [HIV]?

Availability of peer support network.

Community engagement through peer educators was a powerful motivator for many FSWs to seek HIV related information and services. Some participants recounted a story about the feelings of social isolation and loneliness when someone receives the HIV-positive diagnosis. For example, one participant explained that;

Testing HIV-positive makes you feel as if you are left alone, and that everyone leaves you because of your condition

Participants pointed out that, the presence of the Sauti Project and interaction between HIV positive FSWs with peers who had similar situation minimized fear of isolation and revived hopes of living like others even after an HIV-positive diagnosis. Usually, HIV positive FSWs felt less alone and less stigmatised when they met other FSWs with the same HIV status. HIV positive FSWs felt to be part or integrated into the networks of people living with HIV, thus willing to be escorted by their peers to the health facility for treatment and care. A participant during the discussion pointed out that:

The advantage is that the one who tells you go to the health facility has been through the same situation and the same feelings you have, so it becomes easier to accept because you know that whenever you have a problem, she can guide you on what to do. Also, they know each other; encourage one another … so you are not alone anymore

System-level barriers.

Fear of breach of confidentiality: All participants perceived referral to be the services offered to individuals diagnosed with HIV either through peer escort or given self-referral to the nearby health facility for HIV care registration and initiation into HIV care and treatment. Subsequently, in all PGDs, participants reported fear of breach of confidentiality as one of the significant barriers for FSWs diagnosed with HIV to enrol to the health clinic for care and treatment services. FSWs indicated that some healthcare providers working at the clinics “tell other people about their clients’ HIV status”. It was also reported that some of the health workers who were aware of individual positive status “alert their friends not to have sexual relations with the infected FSWs”. Participants in one of the PGD sessions echoed this concern:

Male healthcare providers are leading gossipers … after attending to you [[s]] he will start to caution his close male friends to take care … not to approach us, and later everybody in the village will know it [our HIV status]

Similarly, some participants had mistrust toward some peer educators (PEs) and home-based care (HBC) providers who were instrumental in linking clients to the HIV care clinics. In particular, participants complained that some PEs and HBC providers working with the project disclosed their status, which led to bad naming (called ‘prostitute’) and even discrimination by other community members. Discussing this barrier, some participants pointed out that:

They [PE and HBCs] help us, but some after they have sent you to the health facility you will hear the news about your HIV status everywhere and […] they [community members] will start to refer you as a prostitute

Some when they escort you to the hospital, they end up spreading the information to other people that you have the virus.

Prolonged pre-enrolment period: Participants reported that if one accepted to be linked to HIV care, she was supposed to attend up to three sessions of adherence counselling before enrolment. Most participants felt that attending three days’ counselling sessions alerted other community members about their HIV status. Moreover, attending all those sessions needed transport cost and time that could help with something else. Some participants cited cases where clients had attended one or two sessions and decided to abscond. A remark made by a participant in one PGD session explains this sentiment:

We know of people who were referred to HIV care clinic, but after realising that they will have to attend three days’ training with other HIV-positive people, they decided to abscond […]. Some attend [the training] for one or two days […] you know the nurses don’t tell you when you will start getting the medicine, so you go there the first day, you go to class then you are told to come back next day—there is no clear information

Vertical and integrated HIV service delivery approach: Participants reported that some health facilities had dedicated rooms or blocks to attend for testing and counselling, enrolment to care, and administration of ART. According to participants, people who attend the facility for other general health services could easily see people living with HIV visiting those rooms. Providing services in dedicated venues were perceived to be stressful since it minimises privacy. One FSW who was against offering HIV care and treatment services at dedicated facilities away from other general services said:

You usually see them [PLHA] congested at one place. So when you go there, everybody knows that you have that dangerous disease and you become a talk of the town

Many participants reported that some HIV care clinics were open for 8 hours only (from 8 am to 4 pm), and thus could not attend clients who visited the health facilities after that time. Participants suggested that the availability of health providers at the health facility all the time would help to overcome the challenges of working hours. In some districts, participants complained that some health facilities were enrolling clients to HIV care only for a few dedicated days in a week.

If you go there on days when they do not enrol new people, they look at you and ask you to come another day

Societal-level barriers.

Myths and rumours about ARVs: Many participants discussed widespread concerns about the side effects of ARV drugs, and that these rumours discouraged many FSWs from accepting linkage to care. Participants reported that everyone in the community knows that treatment with ARVs can cause horrifying dreams and loss of appetite. Besides, FSWs thought that ARVs could have adverse outcomes including further weakening for people suffering from opportunistic conditions. In some PGD sessions, participants expressed the belief that treatment with ART was behind the increased spate of sudden deaths occurring in the community.

Many people believe that if you start it [ARV] early it makes you tired and makes your body to weaken quickly. You can even die suddenly even when you are sleeping … You may feel a headache and die suddenly.

FSWs who self-identified themselves to have enrolled in treatment complained that ARVs were very powerful and required one to eat well before taking them. For many participants, the ‘strength of ARVs’ and the ‘need to eat well’ caused fear among FSWs who tested HIV positive but who had not started treatment with ART. Following this belief, participants thought it was wise for FSWs to choose whether to take the drugs despite the anticipated adverse outcomes.

There are times when the family go to bed without food, and now you will know what to do with your medicine … if that is not death

The stigma associated with receiving HIV care: Almost all participants mentioned HIV stigma as a key barrier for FSWs to successful link to care. Some participants reported having experienced stigma, particularly when they interact with health workers. There were, for example, cases where FSWs reported to have been delayed services or called bad names when visiting health facilities. Participants indicated that it was also common for peers and other community members to use derogative terms and naming toward people living with HIV. According to participants, people who utilised HIV care services were referred to using stigmatising and derogatory terms such as; ‘maiti mtarajiwa’ [lit. expectant corpse], ‘maiti anayeishi’ [lit. a living corpse], ‘moto’, [fire] and ‘maiti inayo tembea’ [lit. a walking corpse]. These had a negative connotation on their day-to-day life and their work as FSWs since it could cause them to lose clients.

HIV stigma did not only come from society but also close relatives and family members. Participants provided various examples to explain the family reactions after finding out that FSWs were using HIV medication:

She [FSW] started using the medicine [ARVs] immediately after she tested HIV-positive … but because she believed they [family] would support her, she told her mother that she is using those [ARVs]. Instead of offering support, her mother called her aunt [FSW’s aunt] to tell her that the girl brought bad luck in the family and that she was a prostitute, and everyone who came home was told the same things […] even if it were you, would you accept to start the medicine.

With multiple sources of stigma, participants discussed that some FSWs do not accept to go to HIV care clinics located within their villages. A few discussed that some FSWs avoided using transport offered by Sauti for clients to go to a nearby health facility to be linked to care. They instead travelled to other places to get the medicines and use them secretly.

Individual-level barriers.

Perceived health status: Low HIV risk perception plays a role in FSW’s acceptance of linkage to care. Those who had not developed symptoms of HIV considered themselves as being healthy, thus they did not see the importance of linking to care even when they were told that they were infected. Participants expressed that most FSWs who had taken HIV-test out of curiosity and found to have HIV infection, denied the test results because they felt that they were in ‘good physical health’. In one FGD, a young woman stated that:

You feel well just like everyone … so you continue with your life.

Contrary to popular expressions, however, a few participants felt that these sentiments were a result of limited awareness on the mechanisms through which HIV progresses and the importance of early initiation of care. Participants believed that HIV can be in the human body for several years while multiplying, and that manifestation of the symptoms is an indication that an individual is about to die.

Fear of spoiling the relationship with partners or losing the clients: Some participants reported that FSWs fear to link to care as this would alert their potential clients and they would lose the clients. For those who had permanent partners feared to scare them, this would lead to separation or break the relationships. With this fear, participants reported that most FSWs found to be infected avoided to be linked to HIV care clinics because their HIV status would be known.