Introduction

Global efforts to control HIV have recently focused on high-prevalence geographic areas and populations perceived to be fueling HIV epidemics [1–5]. Fishing communities in sub-Saharan Africa are an example of such settings. They are frequently characterized by unique social and economic factors that contribute to high HIV prevalence, including high rates of transactional sex and sex work [6–8], mobility and migration [9, 10], a “culture of risk denial or risk confrontation” associated with an often dangerous industry [11], social marginalization [9], alcohol use [9, 12, 13], and limited access to health services [7, 9, 11, 12].

HIV self-testing (HIVST) is a promising approach to expanding HIV testing services and may have particular benefit in these unique environments [14], which are often underserved by traditional HIV testing services. HIVST is a process in which a person who wants to know his or her HIV status uses a kit to collect a specimen, performs a test (generally a rapid diagnostic test), and interprets the test results for himself or herself; positive results require confirmation testing at a health facility [14]. However, there has been little research on HIVST in high-prevalence geographic areas or with key populations in sub-Saharan Africa [15]. A recent systematic review of values and preferences related to HIVST conducted to inform World Health Organization guidelines identified 90 peer-reviewed articles worldwide [14]. Twenty-five were conducted in sub-Saharan Africa, but all were limited to general populations [14]. The majority of participants in these studies reported high acceptability of HIVST, but some voiced concerns about misuse [16–20].

Only three studies have explored perceptions of HIVST among sex workers—each from Puerto Rico, China, and Kenya [21–23]. In Puerto Rico, male and transgender female sex workers voiced concerns about confidentiality and triggering violence with clients [21]. In China, sex workers reported high acceptability of oral HIVST kits, but doubted the accuracy of the results [22]. In Kenya, secondary HIVST kit distribution to sexual partners of female sex workers was relatively successful, as 75% of participants distributed self-tests to their primary partners and 81% distributed more than one self-test to clients [23]. Provision of self-tests in this study also resulted in high levels of couples testing and significant changes in reported sexual risk behavior [23]. No studies have been conducted among fisherfolk, despite the high HIV burden among these communities. Furthermore, despite global efforts to focus HIV resources on high-risk populations, we do not know how values and preferences around HIVST may differ among high-risk groups compared to those of the general population.

Fishing communities along Lake Victoria in Rakai, Uganda have high HIV burdens, with a median HIV prevalence of 42% (range 38–43) [12], several times the Ugandan national average of 7.1% [24]. High HIV prevalence in these communities may be related to a combination of unique factors related to the fishing economy, including significant stigma, frequent unprotected sex, risk denial, and substance use [25–27]. Sex work is prevalent and individuals report higher rates of HIV risk behaviors compared to individuals in nearby mainland areas, including having more than one sexual partner in the past year and using alcohol before sex [12]. In this setting, we sought to qualitatively examine values and preferences related to HIVST among community members and health care providers in both mainland and high-risk fishing populations, including with sex workers and fishermen, in Rakai District, Uganda.

Methods

Ethical approval for this study was received from the Western Institutional Review Board (20151088) and the Uganda Virus Research Institute Ethics Committee (GC-127-16-10-525). Prior to enrollment into the study, all recruited individuals provided written informed consent for participation. We conducted in-depth interviews (IDIs) with 12 health care providers (clinicians and HIV counselors) and 21 community members from both high-risk fishing communities and low-risk rural mainland communities in Rakai, Uganda (Table 1). Participants from fishing communities were purposefully selected to include sex workers and fishermen. Additionally, we conducted 6 focus group discussions (FGDs), with a total number of 55 participants, stratified by sex, location, and age (Table 1), to examine group discourse and social norms related to HIVST in both settings. Individuals were recruited through initial seeds provided by Rakai Health Sciences Program (RHSP) community health workers or through participants in the Rakai Community Cohort Study (RCCS) who had agreed to be re-contacted for future studies, with additional snowball sampling from initial participants. We followed an iterative process of data collection and analysis, ending recruitment when we felt data saturation was reached.

Download:

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

Table 1. IDI and FGD participants by location, group, and sex.

https://doi.org/10.1371/journal.pone.0183280.t001

Eligible participants were invited to participate in an interview or focus group at a time and location of their convenience. Before starting data collection, interviewers explained the purpose of the study and obtained written informed consent. Interviews and FGDs were conducted in Luganda or English and audio-recorded with permission from participants. Questions explored perceived positive and negative aspects of HIVST and implementation preferences. Participants were compensated with 10,000 Ugandan shillings for their time and transport costs.

IDIs and FGDs held in Luganda were translated by experienced graduate research assistants into English. All data was transcribed, coded and analyzed using a team-based matrix approach adapted from Framework Analysis [28]. Matrices allowed investigators to summarize and compare the data based on participant characteristics including sex, occupation, age, marital status, and location. Working collaboratively, several team members first read through the transcripts, then discussed and developed a set of codes to be used in the matrices that reflected both the original research questions and emerging themes. We then summarized each participant’s response for each code and included illustrative quotes. The matrices were completed by two team members independently for a set of sample transcripts, then discussed and refined. The finalized matrix was then used to code all other transcripts. In conjunction with the coding process, investigators wrote memos to develop conceptual categories and to track emerging insights on the data.

Results

Table 1 presents the number of IDI and FGD participants by location, group, and sex. We conducted interviews with 10 men, 11 women, and 12 health care providers. Of these, approximately half were in fishing communities, where all men were fishermen or boat owners and all women were sex workers, and half in mainland communities, where occupation was not recorded. Among the participants who chose to self-report their HIV status, there was a mix of HIV-positive and HIV-negative individuals. More female sex workers reported known HIV-positive status than other participants. As there were no major differences between FGD and IDI participant responses, results from both data collection activities are combined in the themes presented below.

Most participants had never heard of HIVST. One mainland woman reported hearing of a school-based self-testing program in Tanzania, where students could test themselves privately and then do a repeat test with a counselor if warranted. Another female sex worker reported she had previously attempted to use an HIVST kit with her boyfriend who obtained it from the army, but due to what she described as mistakes in execution with the buffer, they were not able to obtain results. In this case, the sex worker knew she was HIV-positive and her boyfriend had asked her to test before having sex; when the test failed, she reported that they did not have sex, but she said she would have used it as a way to disclose her status had the test worked.

After being presented with the idea of HIVST, most participants felt it could have significant benefits, although others had serious concerns. These perspectives are outlined below.

Discussion

This qualitative study describes perceptions of HIVST among both lower-risk mainland and high-risk fishing community members in Uganda (including sex workers and fishermen), populations that may benefit from self-testing programs following 2016 World Health Organization (WHO) guidelines recommending HIVST [14]. Participants discussed a range of benefits including privacy and convenience, noted potential barriers related to education and support, and suggested cost and distribution strategies if implemented in their communities. Many of the perceived benefits and barriers described by participants in this study align with findings from other studies on values and preferences related to HIVST globally [14] and among key populations (26). Recent systematic reviews found that barriers to HIVST uptake among key populations include doubts about accuracy, lack of provider assistance, and concerns about use, while facilitators include convenience and privacy [14, 15]. While the overall majority of participants concluded that the benefits of HIVST outweighed the risks, participant responses suggest HIVST may particularly address some of the economic and social barriers to facility-based HIV testing unique to fishing community settings.

Participants’ primary concern about HIVST was the lack of professional support if testing outside of a health care facility. A recent systematic review identified similar concerns about the risks of violence and nondisclosure from studies conducted among general populations in Africa [14], though reports from Nigeria [18], Kenya [19], and Zambia [17] suggest many participants anticipate the benefits of HIVST will outweigh these risks. WHO guidelines state that there is no evidence of any serious adverse events or harm as a result of HIVST, but that inclusion of risk mitigation and reporting strategies is still important [14]. There are several ways education and professional support could be incorporated into a self-testing model to alleviate some of these concerns. As some participants in this study noted, the most conservative option would be to distribute HIVST kits at health facilities where individuals could receive traditional pre-test counseling. However, this strategy may face the same challenges as existing facility-based services. Evidence from pilots in South Africa [29] and Malawi [30] support the feasibility of providing HIVST education outside of facility-based settings. If HIVST kits are available outside of clinics (e.g. at pharmacies, lodges, bars), self-testing information could be communicated by community health workers [20] or incorporated into existing health education campaigns, a strategy supported by the WHO [14]. Effective pre-test counseling in any form could support correct self-testing practices and accurate interpretation of results, emotionally prepare patients to receive results, and emphasize linkage to confirmation testing and treatment. However, this does not eliminate the need to create a more approachable healthcare environment for patients and decrease HIV stigma in these communities.

While most participants wanted to access HIVST kits at health facilities, young men and fishermen favored more convenient points of access such as bars and lodges. HIVST pilots in Kenya [31] and Zimbabwe [32] tested non-facility-based HIVST kit distribution strategies (distribution by sexual partners and community health workers, respectively) and both reported increased rates of kit use and successful testing. Studies from Kenya, South Africa, and the United States suggest the accessibility and privacy of HIVST may be uniquely effective at reaching previously unengaged populations, particularly men and youth [31, 33–35]. Since these groups have high rates of HIV and sexual risk behaviors [12], it may be advantageous to consider their specific preferences to expand coverage of HIV testing beyond existing systems.

The use of HIVST to learn a new partner’s HIV status before engaging in sex appealed to both men and women in this study. Female sex workers, in particular, saw this as a benefit since so many clients refuse condoms. The use of HIVST as a harm reduction strategy among populations who report inconsistent condom use has been explored in several U.S.-based studies [36–38]. Data from Kenya [31] and the United States [36, 39] indicate HIVST may be used as a screening tool for potential partners, and suggest individuals who receive a positive self-test result may be less likely to engage in condomless sex [31, 39]. However, sex workers whose clients receive negative HIVST results may be less likely to use protection with these partners. The 2016 WHO HIVST guidelines conclude the benefits of HIVST outweigh these potential risks, but advise that HIVST programs include clear messaging about potential risks of using HIVST in this way [14].

Participants in this study were not specifically asked about their preferences regarding oral swab testing versus whole-blood finger-prick testing, but several participants expressed concerns about sharp testing materials that could be used as weapons or accidentally infect others when discarded inappropriately. One can assume these risks would not be anticipated with oral swab testing. Other studies have generally suggested that people prefer oral swab to blood-based self-testing [15, 40], usually because it is pain-free [22, 41] and no blood is involved [22]. However, previous studies, including with sex workers, reported that some pilot participants question the reliability of these results (23). While oral specimen-based rapid tests have shown high positive predictive value in high-prevalence settings, they have slightly lower sensitivity and positive predictive value in low-prevalence settings [42]. Previous challenges interpreting rapid diagnostic tests in Rakai [43] suggest additional research may be needed to identify the most effective method of informing testers how to accurately complete and interpret HIVST results. Program implementers will need to consider these trade-offs when developing HIVST programs.

This study had several limitations. Although a few sources report informal availability of HIVST kits in some private clinics in Uganda [44], the majority of our participants reported never having heard of HIVST prior to this study. Interviews thus focused on hypothetical circumstances rather than direct experiences with HIVST, although interviewers did try to probe participants to describe the factors influencing their choices and explain any previous experiences relevant to their responses. Some meaning could have been lost in translating interviews from Luganda to English, although the research team that collected the original data was also involved with translation, interpretation and analysis. Despite these limitations, we believe we reached saturation of themes for our primary research questions and believe the findings of this study may be transferable to similar contexts.

With the new 2016 WHO recommendation for HIVST [14], we anticipate a greater expansion of self-testing programs globally. In Uganda, several HIVST pilot initiatives are underway, including two testing peer-distributed HIVST strategies among key populations–female sex workers in Kampala [45] and fisherfolk in Buliisa [46]—and another testing a partner-distributed strategy among women attending antenatal care facilities in the Central region [47]. The Ugandan Ministry of Health is currently considering introducing oral HIV self-testing kits, based on evidence from these pilot programs, to increase low uptake of HIV testing among men in particular [48]. Evidence from these pilots will continue to inform HIV self-testing scale-up in Uganda and around the world.

Conclusion

Fishing community members often have high HIV prevalence and incidence, yet face financial and social barriers to facility-based services. The privacy and convenience afforded by HIVST may particularly benefit these key populations, although these same features may be problematic when associated with a lack of supportive counseling and linkage to care after an HIV-positive test result. Overall, nearly all participants in this study concluded that the personal benefits of HIVST outweighed the risks, that they would be willing to self-test, and that they would recommend self-testing to others. HIVST may be a unique tool to reach these groups, though program implementers will need to consider how to balance accessibility with necessary professional support.

Acknowledgments

We would like to thank all of the participants who contributed their time and thoughtful responses to this study and the RHSP community health workers who mobilized them. We thank the Rakai Health Sciences Program leadership team, administrative team, and qualitative data collection team who made this study possible.