Acceptability of HIV self-testing to support pre-exposure prophylaxis among female sex workers in Uganda and Zambia: results from two randomized controlled trials

HIV pre-exposure prophylaxis (PrEP) is highly efficacious for the prevention of HIV acquisition in men who have sex with men [1] and heterosexual serodiscordant couples [2, 3]. Although few HIV seroconversions have been documented in the context of PrEP implementation programs in the United States [4–7], there is concern related to the development of resistance for individuals who acquire HIV while taking PrEP [7]. Routine HIV testing is therefore a core component of patient management in PrEP programs. The World Health Organization [8] and the Centers for Disease Control and Prevention [9] recommend quarterly HIV testing for all individuals taking PrEP. However, frequent HIV testing can be burdensome to some PrEP patients, particularly for those who face logistical barriers to accessing healthcare [10].

HIV self-testing allows individuals to test for HIV at the time and place of their choosing, and may reduce some barriers to accessing regular HIV testing. In the context of PrEP implementation programs, HIV self-testing might be a useful complement to clinic-based testing [11]. For example, if the sensitivity and specificity of HIV self-testing is adequate for people taking PrEP [12], HIV self-testing could be used to screen for HIV instead of clinic visits for some time points. HIV self-testing has generally been shown to be acceptable [13, 14] and has been implemented in diverse populations of users [15].

Female sex workers (FSW) in Sub-Saharan Africa are disproportionately affected by the HIV epidemic [16]. PrEP has the potential to substantially alter the course of the HIV epidemic among FSW, particularly as a user-controlled prevention intervention that does not require negotiation with a partner, such as male condom use [17]. FSW often receive a higher price for engaging in condomless sex, and thus may be economically disincentivized from using male condoms in particular [18]. PrEP may therefore be a beneficial tool for FSW in some contexts [19]. A demonstration project of PrEP integration into routine care for FSW in South Africa demonstrated good uptake but suboptimal retention in PrEP care [20]. Barriers to accessing routine HIV testing exist for many populations of FSW, including logistical barriers such as timing of clinic hours and interpersonal barriers such as anticipated stigma from healthcare workers [21–24], which may contribute to poor retention in PrEP care.

Here, we report PrEP acceptability among FSW participating in two trials of oral HIV self-testing delivery models in Zambia [25] and Uganda [26], and the acceptability of HIV self-testing in the context of PrEP use.

FSW in both Zambian transit towns and urban Uganda reported high interest in PrEP and preferred HIV self-testing over standard HIV testing services at clinics to test for HIV regularly (every 3 months) while on PrEP. HIV self-testing for regular, repeat testing was acceptable among FSWs in this study, especially among those who previously had the opportunity to self-test. FSW who had the opportunity to self-test previously may have preferred this approach over standard HIV testing services because they were less intimidated by the new testing technology and had a favorable experience HIV self-testing. These results indicate that HIV self-testing could be used to support PrEP delivery by moving some of the burden of HIV testing outside of health clinics, preventing stigma and discrimination of FSW from healthcare providers and maintaining the confidentiality of FSW taking PrEP [22].

Daily oral PrEP has been shown to be acceptable [29] and implementation to be feasible [20] among other populations of FSWs. In the current study, acceptability of PrEP for prevention of HIV acquisition was high among FSWs in both Zambia and Uganda, but some PrEP modalities were more popular than others. In both study settings, FSWs preferred daily oral PrEP or quarterly injectable PrEP over PrEP in the form of a vaginal gel or ring. This preference is likely explained by the prevalence of various types of non-barrier contraceptive methods available to women in these countries. A recent study conducted among the same population of FSW in Zambia found that the majority of FSW who used a method of non-barrier contraception used injectable contraception; the second most common form of non-barrier contraception was the oral birth control pill [30]. Less than 1 % of Zambian FSW in that study reported use of the vaginal ring [30]. Lack of familiarity with contraceptives inserted in the vagina might have made FSW less interested in PrEP modalities that utilized this strategy.

Acceptability of HIV self-testing was extremely high across randomization arms in both Zambia and Uganda. More than two thirds of participants in the standard of care arm preferred HIV self-testing to standard HIV testing services while taking PrEP in both settings, despite never having had the opportunity to self-test themselves. Willingness to test for HIV every three months was also near universal in both settings, thus suggesting that barriers to HIV self-testing among members of this population are likely attributable to logistics or stigma and discrimination from healthcare providers rather than a lack of will. As demonstrated in the two parent trials, HIV self-testing has the potential to help FSW overcome some of these barriers and achieve both high HIV testing coverage and high levels of repeat testing [22, 26]. A recent study, however, suggests that FSW might have difficulty interpreting HIV self-test results [31], indicating that appropriate pre-test training and on demand support should be considered with the implementation of HIV self-testing.

This study has both strengths and limitations. We found consistent results in two diverse populations of FSWs, improving the generalizability of our findings. Participants were engaged in a randomized trial of HIV self-testing, and thus we were able to compare preferences for HIV self-testing between women who had been randomized to receive a self-test compared to those who did not receive a self-test. Questions related to acceptability of HIV self-testing were therefore not hypothetical for women who had previously had exposure to an HIV self-test. None of the participants in this study, however, actually used or had access to PrEP, and thus they might not have fully understood the research assistants’ explanation of this HIV prevention intervention. Also, as a hypothetical scenario, participants may be more or less likely to indicate that they would use an intervention than they would if they were actually making a decision in real life. Actual behaviors may therefore differ from anticipated behavior. Finally, our selection of study participants may have induced bias in the effect estimates because we selected on a variable (i.e., HIV-negative status) that was reported after randomization. However, PrEP is only indicated for individuals who test negative for HIV, and participants who are living with HIV may have substantially different responses to hypothetical questions regarding HIV prevention. Furthermore, to induce bias, the study randomization arm would have to causally affect HIV status. It is unlikely for this to occur in the one-month timeframe of the study.

In many sub-Saharan African countries, FSW are a priority population for HIV prevention interventions because of their increased risk of both HIV acquisition and transmission [16, 32]. As more sub-Saharan African countries, including Kenya, Uganda, and South Africa, begin to scale PrEP nationally, understanding preferences for PrEP and interventions for ongoing engagement in care is of increased importance. Here, we found that PrEP is highly acceptable among members of this population and that HIV self-testing is preferred over standard HIV testing services to support regular HIV testing while on PrEP. Previous studies have found HIV self-testing to be effective at increasing frequent testing among FSWs [25, 26]. Policy makers should consider both scaling PrEP to FSW living in high prevalence areas and using HIV self-tests to support PrEP delivery and facilitate detection of breakthrough HIV infections.