South Africa is home to the highest number of PLWH¹ and KwaZulu–Natal (KZN) province has the highest HIV prevalence nationally (1)². South Africa's HIV testing and care cascade stands at 92-70-64³, highlighting the great progress made in terms of people knowing their HIV status, but more effort is still needed to reach the United Nations Programme on HIV/AIDS (UNAIDS) 95-95-95 targets by 2025⁴. Individuals younger than 35 and men account for most of the people untested for HIV in KZN (2). Inconvenient clinic operating times, stigma, and preference for using traditional medicine are some of the barriers which prevent men from accessing facility-based HIV testing services (HTS) (3, 4). Alternative strategies to facility-based HTS are needed to reach these hard-to-reach populations.

HIV self-testing (HIVST), using a simple oral-fluid or blood-based self-test at a time and place convenient to the person testing, is a promising alternative strategy that addresses many barriers associated with facility-based HTS and has the potential to increase testing among persons who do not usually go to clinics (5). Different HIVST kit distribution strategies such as distributing through peer adolescent networks (6) and distributing to female sex workers (7) can be used to reach vulnerable populations who may experience barriers to accessing facility-based testing.

Secondary distribution of HIVST, where an HIVST kit is given to an individual for distribution to a third party, has been shown to reach male partners of women attending health facilities and can facilitate couples-testing (8, 9). Secondary HIVST kit distribution by peers in fishing communities was successful at finding untested men in one study in Uganda (10). In Sub-Saharan Africa, secondary kit distribution studies have primarily engaged women to distribute HIVST to partners and contacts (8, 9, 11). Little is known about secondary HIVST kit distribution using index distributors of both male and female PLWH as distributors in South Africa. We aimed to assess the willingness of both male and female PLWH to distribute HIVST kits to social networks including sexual partners, and to determine kit usage, test positivity, and linkage outcomes from this distribution approach.

The HIVST kit distributors provided written informed consent to participate in the DO ART study and a separate consent to receive and distribute HIVST kits. Informed consent was obtained from secondary kit recipients over the phone once they called to give results after using the HIVST kit. Consent to use the test was implied if kit recipients contacted the study to report their results. This study received ethical approval from the Human Sciences Research Council, the University of KwaZulu-Natal and the University of Washington's Ethical Committees.

A total of 63 HIVST kit distributors took 218 HIVST kits to distribute to secondary contacts with a median of 4 kits (3 blood and 1 oral based) taken per participant. The HIVST kit distributors dispersing kits were 48% (n = 30) male and 52% (n = 33) female. The HIVST kit distributors median age was 34 years (26–42.5). Fifty-five percent (n = 35) indicated they were in a relationship and 84% (n = 52) indicated having 1 sexual partner (Table 1). Seventy five percent (n = 47) of those who took kits for distribution were unemployed; 89% (n = 54) reported an educational level at secondary or above. Seventy six percent (n = 48) indicated not knowing their partner's HIV status and 46 (73%) kits were reported to be intended to be given to a friend; 13 (6%) kits were intended to given to a sexual partner.

A total of 147/218 (67%) intended kit recipients contacted the study after using the kits. Recipients were [65 (44%)] males and [82 (56%)] females. Eighty-five recipients (58%) reported having previously tested for HIV prior to using the HIVST kit, and 143/147 (97%) reported their HIVST results. Four (3%) participants had missing data and therefore were recorded as not reporting their results (Table 2). Most of the HIVST results reported were negative at 92% (n = 132) (Figure 1). From the 11 (8%) recipients who reported positive HIVST results, only 1 (9%) reported being linked to care.

In this evaluation of secondary distribution of HIVST kits by male and female PLWH, we found that secondary kit distribution achieved a “return rate” of 143 of 218 (66%) kits used with results reported. The outcome of 75 (34%) kits remained unknown. From results reported, 62/147 (42%) indicated being first time testers, highlighting the potential secondary HIVT kit distribution has to find first-time testers. Even though 11 (8%) kit recipients reported to have tested positive, only 1 (9%) linked to care, leaving 10 (91%) with unknown linkage. This highlights the fact that HIVST can get people tested, but more support is needed to fully link them to care.

In finding first-time testers, our results are similar to findings by Choko et al. (10) who reported that secondary kit distribution using fishermen peers reached 23/89 (26%) men who had not previously tested for HIV in Uganda. Finding first-time testers will be an important step to overcome if the UNAIDS first “95” testing cascade will be achieved in KZN by 2025. Secondary kit distribution could be used an alternative strategy to find those not previously tested.

One challenge in evaluating secondary kit distribution is the reliance on testers to self-report their results, thus the number of HIVST results we received is an underestimate of the number of tests actually used. In their study of secondary kit distribution, Choko et al. (10) found that 87/94 (86%) participants recruited through fishermen peers in Uganda, reported their HIVST results. This reported percentage is much higher than the 66% reporting we found, but a major difference between the two studies is that the peers who distributed the kits in Uganda had to bring back the kits irrespective if they had been used. In contrast Hensen et al. (13) reported that 460/825 (56%) kit recipients confirmed to have used the kits after secondary distribution by community HIV care providers (CHIPs) in 4 communities within Zambia. In this study, kit recipients also had to return the kits to the CHIPs whether used or not. Across all studies, more kits were used and had results reported compared to unused/unknown kits, regardless of the reporting approach. We recommend that HIVST through secondary kit distribution is sufficiently high yield that it should be scaled up, even with the understanding that not all kits may be used or lead to immediate engagement in care.

From the HIVST kit distributors who dispersed kits, 48 (76%) indicated not knowing their partners' HIV status, yet only 13 (6%) kits were intended to distributed to a sexual partner. These results contrast with findings from Masters et al., who reported that secondary kit distribution by women seeking antenatal and postpartum care in Kenya promoted partner testing (8). Potential reasons for partners not wanting to distribute kits to partners could be that positive results can invite stigma, abandonment, discrimination, violence, and emotional distress (14). Secondary distribution of HIVST kits by PLWH to their sexual partners may result in disclosure of the distributor's HIV status. Additional elements could be incorporated into a program of HIVST distribution to social contacts in order to minimize social harms associated with status disclosure and promote partner kit distribution. Thirumurthy et al. (9) described an approach where secondary kit distributors were taught to use discretion when introducing self-tests to sexual partners by assessing likely reactions of their partner(s) as well as the risk of intimate partner violence. Scale-up of secondary kit distribution in the region should explore using these strategies.

An overwhelming majority of those who reported testing positive did not link to care. Future kit distribution programs should strengthen follow up and linkage to encourage those who test positive to easily link to care. Future research should explore reasons why people who reported to test positive through secondary HIVST kit distribution do not link to care.

This study had several limitations, including that we had no way to contact the kit recipients. We could not confirm if the HIVST kit distributors' intended kit recipients were the ones who received the kits. Also, even though two thirds (66%) of the participants responded with their results, this percentage is lower than the 80% responder percentage reported by a concurrent study providing primary distribution of HIVST kits to recipients in the same areas (5). Some kits may have been discarded before they were distributed, or recipients may not have used their kits. We were unable to ascertain results of kits that were used, or characteristics of the person who used the kit, unless the recipient elected to contact the study. Finally, self-reported HIVST results and linkage data are subject to some bias. If kit recipients already knew their status but had not disclosed to the distributor, this could have led to them accepting a test but not using it or not reporting their results.

The strength of the study is that it was done in an operationally friendly way using an existing structure of a community-based study, so it didn't have to create an entirely new infrastructure. The limited infrastructure requirements of this project suggest that results would be similar if translated into a public health setting, for example, implementing a secondary distribution programme in existing HIV treatment sites.

We found that secondary kit distribution using both male and female PLWH as HIVST kit distributors can engage social contacts into HIV testing. However, low subsequent linkage to care among persons testing positive indicates further support with linkage is needed. This strategy can extend testing from facility-based into the community and reach first-time testers, and further efforts to improve linkage should be explored for persons accessing HIVST.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

This study received ethical approval from the Human Sciences Research Council, the University of KwaZulu-Natal and the University of Washington's Ethical Committees. The patients/participants provided their written informed consent to participate in this study.

AS, CC, and RB designed the study. NS, OK, KS, MK, and AV conducted the project and collected the data. NS, AS, and TS analyzed the data. NS and AS drafted the brief report. All authors contributed to the revisions and content of the final manuscript.

This work was generously funded by the Bill and Melinda Gates Foundation (BMGF #OPP1134599).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.