Despite major gains in HIV prevention, treatment and care, uptake among people who use drugs (PWUD) remains inequitable [1]. HIV testing among at-risk groups is central to HIV prevention efforts, in particular as the ‘treatment as prevention’ paradigm has been shown to improve health outcomes and longevity among those infected, to prevent onward HIV transmission, and to reduce healthcare costs [2–5]. However, PWUD are more likely to receive HIV testing later in the course of infection and some only after developing HIV-related opportunistic infections [6]. This disparity to accessing conventional HIV testing services among PWUD reflects the structural, economic, social, and individual-level barriers experienced by this population [6], including perceived and enacted stigma by health care providers [7].

The recent availability of rapid HIV point-of-care testing (POCT) offers new strategies for improving the uptake of testing among PWUD in both clinical and community-based settings [8,9]. Rapid HIV POCT can reduce anxiety associated with waiting for test results: non-reactive results can be provided at the time of testing, while participants with reactive results can receive on-the-spot linkages to confirmatory testing, support, and treatment [10]. Community-based POCT offers the additional advantage of reaching high risk populations who might not otherwise seek care [11], especially if peers are incorporated into service delivery [6]. Training peer leaders who use (and continue to use) drugs as HIV educators has been shown to be an effective mechanism for HIV prevention among PWUD [12–16]. Furthermore, peer-delivery of testing has been delivered within communities of men who have sex with men (MSM) [11,17,18] and among PWUD with high levels of satisfaction and comfort with confidentiality [19].

The city of Ottawa, Ontario has a large network of people who use both injection and non-injection drugs for a city of its size [20–22]. Despite declines in HIV incidence among the general population, HIV rates among people who use injection drugs in Ottawa remain disproportionately high, with an estimated prevalence of 10% [21]. As 43% of PWUD who were diagnosed with HIV report sharing drug use equipment [21], ongoing transmission remains a significant concern. The PROUD Study is a community-based research project that arose in response to the high rates of HIV among PWUD in Ottawa. In this study, we report a novel community-based peer-administered POCT program developed in collaboration with the local public health authority and communities of PWUD. To the best of our knowledge, and with the exception of the integration of peer educators and testers in a pilot project in Vancouver’s Downtown Eastside [19], there are no other examples of engaging PWUD as testers in community-based HIV testing. Our objective was to elicit the factors associated with the uptake of community-based HIV POCT and to inform models of peer-administered testing among this population.

593 participants were recruited for this analysis: 5 (0.8%) were excluded due to the unavailability of POCT and onsite nursing on the day of their interview and 38 (6.4%) were excluded because of self-reported HIV infection. Of those remaining, 550 (92.7%) were offered HIV POCT: Peers and medical student/allies offered testing to 233 (42.4%) and 317 (57.6%) participants respectively. There was no significant difference in uptake of testing between those offered a peer vs. a non-peer interviewer (83.7% vs. 83.0% respectively, OR = 1.05; 95% CI = 0.67–1.66). The most common reasons for testing refusal were no specific reason provided (n = 21, 22.8%) and recent HIV testing (n = 20, 21.7%). Among participants who consented to POCT, 74 (16.2%) reported never having previous HIV testing.

Table 1 shows the demographic and drug use characteristics of the participants by uptake of POCT. Participants who agreed to POCT were older (43 vs. 39 years, p = 0.02) and less likely to be female (OR = 0.59; 95% CI = 0.35–0.98). Participants who reported ever injecting drugs were more likely to have consented to testing (OR = 1.71 95% CI = 1.09–2.69), but there were no other significant differences in drug use characteristics between the two groups.

Table 2 presents health status and access to health and harm reduction services among those who agreed and did not agree to HIV POCT. No significant differences were found between the two groups.

Table 3 shows variables related to structural risks including associations with the law, sex work, and housing status. Participants who reported having engaged in sex work in the past 12 months were significantly less likely to agree to HIV POCT (OR = 0.48; 95% CI = 0.25–0.92), whereas participants who reported having spent overnight or longer in jail in the past 12 months or to have consumed drugs in jail were significantly more likely to agree to testing (OR = 1.71; 95% CI = 1.05 to 2.78) and (OR = 2.43; 95% CI = 1.07–5.52), respectively.

Only one POCT resulted as reactive and was subsequently confirmed as HIV positive with confirmatory serology.

This study demonstrates the high uptake of a community-based HIV POCT model for PWUD. In order to optimally reach people who use drugs with HIV prevention efforts, our findings support the expansion of novel community-based methods for providing testing, including peer-administered approaches. To the best of our knowledge, and with the exception of the integration of peer educators and testers in a pilot project in Vancouver’s Downtown Eastside [19], there are no other examples of engaging PWUD as testers in community-based HIV testing.

We found no difference in uptake of testing when HIV POCT was offered by a peer versus a non-peer interviewer, supporting previous work demonstrating that people at risk may be comfortable with having POCT administered by peers in a community setting. As has been demonstrated among MSM communities [11], peer testers in our study were able to handle tests, deliver non-reactive results, and provide pre- and post-test counseling in a manner that resonated with participants. Onsite nursing supervision provided peers with support for all aspects of HIV POCT, including: prompting on the POCT process; completing documentation; and guidance to identify when nurse involvement was appropriate. Furthermore, we successfully reached participants who had not previously been tested for HIV, thus providing new opportunities for HIV prevention education and counseling. Previous research has demonstrated that peer-delivered messaging has the potential to increase the interpretation and uptake of HIV prevention information by PWUD, providing value above POCT alone [25,26].

While there were few differences in characteristics among those who agreed and did not agree to HIV POCT, a few key findings emerged. There was high uptake of our community-based testing model among PWUD at high risk of HIV infection, including those who use injection drugs [27] and those reporting drug use while incarcerated [28]. These findings corroborate those of a Baltimore study that found high opt-in rates for onsite rapid HIV testing at probation and parole offices [10]. In our study, testing followed an hour-long interview, and honoraria were provided regardless of whether or not testing occurred, thus testing rates likely reflect a strong desire to be tested among individuals who have previously been incarcerated. However, low agreement for testing among female participants and those involved in sex work suggest the need for more targeted approaches to reach populations who may face multiple sources of stigma and experience increased barriers to accessing HIV testing and treatment services [29]. More specialized approaches, including female only services [29,30], should be explored to better reach these groups.

There are several limitations to our study. First, we used purposive sampling, which allowed us to target recruitment towards those facing multiple and intersecting oppressions among the study participants. However, our findings may not be applicable to all populations of PWUD. Second, characteristics are based on self-report to highly sensitive questions, which may have contributed to social desirability bias and an underreporting of high-risk practices. Third, it is possible that those who declined testing did so not because of the absence or presence of characteristics found to be associated with uptake of testing, but because they were aware of their positive status or had very recent testing, which was not fully elicited by our study. It is also possible that characteristics of testers other than peer or student status influenced uptake. For example, peers tended to be older and male, whereas medical students were younger and female. Participants were provided an option for who would conduct the test when possible, but availability may have also contributed to the ultimate uptake of HIV POCT. Finally, only one new reactive POCT was determined among participants. It is possible that our recruited population was already served by other testing and care services, or that HIV rates may be higher among those who did not agree to testing.

As we move towards scaling up HIV testing opportunities among at-risk groups, there is a need to develop novel models of community-based testing that draw on the strengths of peer involvement to reach individuals who may not seek testing and treatment in conventional health care settings. The PROUD study has shown high uptake of community-based and peer-administered HIV POCT among PWUD at highest risk of HIV acquisition. Future community and peer-based approaches should be implemented towards specific at-risk communities, including women and those involved in sex work, who may face challenges in accessing conventional clinic-based HIV-testing and treatment services.