Prevalence of HIV and sexually transmitted and blood-borne infections, and related preventive and risk behaviours, among gay, bisexual and other men who have sex with men in Montreal, Toronto and Vancouver: results from the Engage Study

Objectives The last Canadian biobehavioural surveillance study of HIV and other sexually transmitted and blood-borne infections (STBBI) among gay, bisexual and other men who have sex with men (GBM) was conducted in 2010. We designed a study to measure STBBI prevalence among GBM in metropolitan Montreal, Toronto and Vancouver and to document related preventive and risk behaviours. Methods The Engage Cohort Study used respondent-driven sampling (RDS) to recruit GBM who reported sex with another man in the past 6 months. At baseline, we examined recruitment characteristics of the samples, and the RDS-II-adjusted distributions of socio-demographics, laboratory-confirmed HIV and other STBBI prevalence, and related behaviours, with a focus on univariate differences among cities. Results A total of 2449 GBM were recruited from February 2017 to August 2019. HIV prevalence was lower in Montreal (14.2%) than in Toronto (22.2%) or Vancouver (20.4%). History of syphilis infection was similar across cities (14–16%). Vancouver had more HIV-negative/unknown participants who reported never being HIV tested (18.6%) than Toronto (12.9%) or Montreal (11.5%). Both Montreal (74.9%) and Vancouver (78.8%) had higher proportions of men who tested for another STBBI in the past 6 months than Toronto (67.4%). Vancouver had a higher proportion of men who used pre-exposure prophylaxis (PrEP) in the past 6 months (18.9%) than Toronto (11.1%) or Montreal (9.6%). Conclusion The three largest cities of Canada differed in HIV prevalence, STBBI testing and PrEP use among GBM. Our findings also suggest the need for scale-up of both PrEP and STI testing among GBM in Canada.


Introduction
The last national Canadian biobehavioural STBBI surveillance study of gay, bisexual and other men who have sex with men (GBM), M-Track, was in 2010 (Public Health Agency of Canada [PHAC], 2011). Since that multi-city surveillance study, most biobehavioural Canadian studies focused on GBM have not incorporated cross-city comparisons. As health care in Canada falls under provincial jurisdiction, the prevalence of sexually transmitted and blood-borne infections (STBBI) and related behaviours may differ among provinces with distinct policy environments.
Half of reported HIV diagnoses continue to be among GBM in Canada (Bourgeois et al., 2017). GBM are 131 times more likely to contract HIV than other Canadian men (Yang et al., 2016). HIV prevalence among GBM tends to be highest in the largest cities (e.g., BC Centre for Disease Control, 2019; Blouin et al., 2019;Ontario HIV Epidemiology and Surveillance Initiative, 2018). Similarly, diagnoses of other STBBIs, such as syphilis, chlamydia, gonorrhea, and hepatitis C virus (HCV), remain common among GBM, with GBM comprising the majority of syphilis cases (PHAC, 2019).
UNAIDS behavioural indicators for Global AIDS Monitoring to reduce the burden of HIV and other STBBIs for high-risk populations, such as GBM, include testing for HIV, testing for other STBBIs, using pre-exposure prophylaxis (PrEP) among HIV-negative persons, and reducing condomless anal sex (CAS) with serodiscordant partners (UNAIDS, 2017). In M-Track, 86.2% of GBM reported they had ever tested for HIV, with proportions for other STBBIs being lower, at 62.6-66.9% (PHAC, 2011). No data were available for PrEP as M-Track predated the 2016 Health Canada approval of emtricitabine-tenofovir disoproxil fumarate for use as PrEP (Government of Canada, 2019). In M-Track, 55% of GBM reported CAS in the past 6 months. Contemporary data are needed as studies across Western countries suggest increases in CAS over time Holt et al., 2018). Temporal increases in CAS may have taken place quickly in Canada, as Lachowsky et al. (2016) found a significant decrease in condom use at last anal sex act in a 2-year period from 2012 to 2014.
The Canadian government has committed to an integrated focus on ongoing surveillance, testing, prevention, and treatment for STBBIs among those most exposed to infections, including GBM (PHAC, 2019). This approach includes promoting sexual health education, condom use and PrEP. Although there are many similarities across provinces, such as efforts to facilitate early diagnosis and initiation of HIV treatment (Ministère de la Santé et des Services Sociaux, 2020; Montaner et al., 2014;Ontario Clinical Care Guidelines, 2020), there are also significant differences in STBBI policies and programs. Some differences include the early implementation of HIV test-and-treat programs as an official policy in BC since 2010 (Montaner et al., 2014), the relatively early listing of PrEP on the Ontario public formulary in September 2017, and the Montreal community-based HIV testing with peers as HIV counsellors since 2009 (Otis et al., 2016). There are also provincial policy differences in the costs of PrEP. In Québec, PrEP has been offered at low cost, as part of the provincial drug plan (Ministère de la Santé et des Services Sociaux du Québec, 2016). In Ontario, while there is partial public coverage through provincial drug assistance programs (Ontario Ministry of Health and Long-Term Care, 2016), most PrEP is still covered by private insurers (Tan et al., 2020). By contrast, in BC, PrEP has been free of charge for all individuals who meet provincial clinical guidelines since January 2018 (BC Centre for Excellence in HIV/AIDS, 2020).
These provincial differences may result in differences in STBBI prevalence and related behaviours among urban Canadian GBM. In addition, the absence of recent STBBI biobehavioural surveillance limits the ability of policy makers to determine the policies and programs that would have the greatest impact on STBBI control in Canada's three largest urban centres. Our study provides updated prevalence estimates for HIV and other STBBIs, and related preventive and risk behaviours among GBM. This study of GBM in Montreal, Toronto and Vancouver is called the Engage Cohort Study (Engage). This paper's objective is to present the study and baseline descriptive statistics on STBBI prevalence and related behaviours as per UNAIDS indicators (2017).

Methods
Engage combines data from computer-assisted selfinterviewing (CASI) and the detection of HIV and other selected STBBIs using biological samples. Engage also involves embedded qualitative studies related to GBM sexual and mental health, blood donation policy and COVID-19 (Grace et al., 2019). Engage was initially designed with a target sample size of N = 2160 (n = 720 in each city). The Montreal site received supplementary funds from the Québec government to increase their sample size to 1200 participants. Using prevalence data from Vancouver (proportion of CAS with a discordant or unknown HIV status partner among participants in Vancouver was 40% per year; Moore et al., 2016), and assuming a design effect of 2 for the RDS sample (Salganik, 2006), our target sample size of 2160 had 80% power to detect a difference of 6-9% in CAS with a serodiscordant or unknown HIV status partner between cities.

Community engagement approach
Before initiating recruitment, investigators at each site consulted with Community Engagement Committees (CECs) comprising staff from local community-based organizations focused on GBM health promotion, including AIDS service organizations, and diverse GBM community members in each city. The CECs provided guidance on the content of the questionnaire, recruitment materials and how the study team could effectively communicate findings to GBM communities.

Sample
Participant eligibility criteria included (1) aged ≥16 years, (2) self-identifying as a man (cisgender or transgender), (3) able to read English or French, (4) live in the metropolitan area of the data collection city, (5) willing to provide biological samples for STBBI testing, and (6) engaged in sexual activity with another man in the 6 months prior to study visit.

Recruitment via RDS
We used respondent-driven sampling (RDS) to recruit participants in Engage. RDS is a modified form of chain referral sampling designed to approximate probabilistic samples by adjusting for selection bias (Heckathorn, 2002). RDS starts with non-randomly selected initial participants (called "seed participants") who then continue recruiting eligible members from their social networks until the target sample size is reached.
The study protocol dictated that all sites initiate recruitment with up to 30 seed participants. Monitoring recruitment progress allowed the team to assess the need to add new seeds. The CECs at each site helped to identify initial seed participants from diverse racial/ethnic backgrounds, gender identities, HIV statuses, risk factors for HIV and other STBBIs, and ages. The CECs were the primary source of recommendations for potential seed participants. In Toronto and Vancouver, advertisements on mobile phone applications, such as Grindr, as well as Facebook and Craigslist, were also used to raise awareness of the study and to reach diverse potential seed participants . After providing written informed consent, seed participants and subsequent participants completed the study questionnaire and biological sampling, and were briefly educated on how to refer other eligible GBM. Each participant received six vouchers to recruit GBM from their social or sexual networks. Participants received CAD$50 compensation for their participation and a secondary incentive of $15 for each eligible GBM they recruited. Each voucher had a unique number to link a participant to his recruiter (thereby establishing recruitment chains and allowing for RDS statistical adjustments) and to facilitate the secondary incentive payment. As per previous RDS literature, we emphasized to seed participants the voluntary nature of recruitment and participation and excluded anyone who indicated that they were coerced into the study. The study was approved by research ethics boards at Ryerson University, University of

Measures
Biological sampling and STBBI testing Participants provide a venous blood sample permitting serological testing for HIV, hepatitis C virus (HCV), hepatitis B virus (HBV) and syphilis; these tests were done according to provincial laboratory algorithms, which are similar in the three cities.
Infection with HIV was ascertained using 4th generation testing (detection of HIV antibodies and p24 antigen); reactive results were confirmed using complementary testing (e.g., Western blot analysis). Similar testing was done to confirm HIV infection for participants known to be living with HIV. In Vancouver, participants known to be living with HIV were also offered the option of either confirming their diagnosis with a point of care (INSTI®) test, or by requesting confirmation from their primary care physician.
HCV infection was determined based on a positive HCV antibody result. For men reporting or found to have a current or past HCV infection, HCV RNA testing was also done to evaluate active HCV infection. HBV status was ascertained using testing for hepatitis B surface antigen (HBsAg). We defined active HBV infection as HBsAg positivity.
Study participants also provided urine, pharyngeal swabs and rectal swabs. Screening for Chlamydia trachomatis and Neisseria gonorrhoeae was done using nucleic acid amplification testing (NAAT) or culture, based on provincial laboratory testing procedures. Due to provincial differences, for throat and rectal samples, half of the participants in Toronto had a culture and half had a NAAT, whereas in Montreal and Vancouver, all specimens underwent NAAT testing. For gonorrhea and chlamydia, any positive result on a urine, pharyngeal or rectal specimen was coded as a detected infection.
Test results were made available to study participants within 2 weeks after collection. Staff provided all participants who tested newly positive for HIV or other STBBIs with linkage to local care and treatment providers. The Vancouver site provided treatment for gonorrhea, chlamydia and syphilis on site. Study staff also provided STBBI transmission risk reduction counselling.

Questionnaire
We used the Sexual Health Framework (Ivankovich et al., 2013) and the Global AIDS Monitoring preventive and risk behaviour indicators (UNAIDS, 2017) to focus the questionnaire on the societal and community contexts, social relationships, and individual characteristics of sexual health among GBM. For the present analysis, we examined four preventive and risk behaviours: (1) HIV testing, (2) other STBBI testing, (3) PrEP use (continuous or on-demand) in the past 6 months, and (4) engagement in serodiscordant condomless anal sex (SDCAS) at least once in the past 6 months.

Data analysis
We conducted a cross-sectional analysis of the baseline data collected at study enrolment. We used the RDS-II estimator (Volz & Heckathorn, 2008) for all analyses. Network size was defined as the number of eligible GBM the participant knew who lived or worked in his city (Montreal/Toronto/ Vancouver). Examples given to participants included GBM whom the participant sees or speaks with regularly, such as close friends, regular sex partners, and roommates. For the lower limit, we set the minimum value to 1 as participants had to be sexually active with another man in the last 6 months to be eligible for the study. We selected an upper limit of 150, following estimates on the maximum number of possible current relationships (Dunbar, 2010).
For this analysis, we calculated crude point estimates and RDS-adjusted local population-based estimates with 95% confidence intervals (CI) for demographic characteristics, STBBI prevalence, test results and STBBI preventive and risk behaviours. We also examined differences in the RDS-adjusted estimates using non-parametric tests considering unequal variance and sample size across cities. RDS recruitment data were analyzed using RDS Analysis Tool version 7.1 (Volz et al., 2012) and statistical analyses were performed using STATA/SE 13.1 (Statacorp LLC, College Station, Texas, USA).

Sample accrual
A total of 27 seeds were recruited in Montreal, 96 seeds in Toronto and 117 seeds in Vancouver. Table 1 Table 2 presents the crude estimates and the RDS-adjusted estimates of demographic characteristics in each city. We observed differences across cities in demographic characteristics. Montreal GBM were older (RDS-adjusted mean age = 37.9 ± 14.6 [SD]) compared with Toronto GBM (RDS-adjusted mean age = 34.8 ± 12.9; p < 0.001) and Vancouver GBM (RDS-adjusted mean age = 35.4 ± 12.6; p < 0.001). Toronto GBM were younger than Vancouver GBM (p = 0.02). Montreal had a higher proportion of GBM with a high school or less education (25.4%) than Toronto (19.5%, p = 0.008) or Vancouver (19.7%, p = 0.004). Montreal had a greater proportion of GBM with ≤$30,000 annual income (66.8%) compared with Toronto (57.4%, p < 0.001) or Vancouver (61.3%, p = 0.01). Table 3 presents the crude estimates and the RDS-adjusted estimates of STBBI testing history and prevalence in each city. HIV prevalence was lower in Montreal (RDS-adjusted percentage = 14.2%) than in Toronto (22.2%, p < 0.001) or Vancouver (20.4%, p < 0.001). History of syphilis infection ranged from 14.4% to 15.9% across cities and there were no significant differences (p > 0.05). In Vancouver, the rectal swab gonorrhea and chlamydia rectal results were not available for 20% of Vancouver participants. Vancouver GBM had a lower prevalence of detected gonorrhoea at any anatomic site (2.1%) than did Toronto (9.1%, p < 0.001) or Montreal GBM (5.5%, p = 0.001). Chlamydia prevalence ranged between 3.1% and 5.8%, and Vancouver GBM had a higher prevalence of detected chlamydia at any site (5.8%) than Montreal GBM (3.1%, p = 0.006). Hepatitis C antibody seroprevalence was greater among Montreal (7.3%) than Toronto GBM (3.9%, p = 0.008) but not more than Vancouver GBM (6.0%, p = 0.27); the difference between Toronto and Vancouver was not significant. The prevalence of hepatitis B surface antigen positivity was greater in Toronto (1.3%) than Vancouver GBM (0.3%, p = 0.04), but not more than Montreal GBM (0.6%, p = 0.36).

STBBI preventive and risk behaviours by city
Regarding HIV testing among self-reported HIV-negative/ unknown participants, Vancouver GBM had the highest proportion of participants (18.6%) who reported never being tested compared with Toronto (12.9%, p = 0.01) and Montreal GBM (11.5%, p < 0.001). Among participants who reported being tested at least once in their lifetime, Vancouver GBM also had the highest proportion of participants (61.9%) who reported being tested in the past 6 months compared with Toronto (54.3%, p = 0.02) and Montreal GBM (52.5%, p < 0.001). Montreal (17.9%, p = 0.003) and Toronto GBM (17.5%, p = 0.02) had a greater proportion of participants who had never been tested for an STBBI other than HIV compared with Vancouver GBM (12.9%). Montreal (74.9%, p = 0.002) and Vancouver GBM (78.8%, p < 0.001) were more likely to have had participants who had been tested for an STI within the past 6 months compared with Toronto GBM (67.4%).

Discussion
In this first cross-city biobehavioural comparison of GBM in Canada in over 10 years, we found differences in HIV and other STBBI prevalence, and in related preventive and risk behaviours. HIV prevalence was significantly lower (14.2%) in Montreal than in Toronto (22.2%) or Vancouver (20.4%). Vancouver GBM were more likely to have been tested for HIV within the past 6 months, but were also more likely to report never having been tested for HIV compared with GBM in Montreal and Toronto. Vancouver's higher proportion of never-tested GBM was unexpected given efforts such as the STOP AIDS initiative and Treatment as Prevention® since 2010 (e.g., Montaner et al., 2014). These findings suggest the need to intensify HIV testing programs for the minority of GBM who have not accessed HIV testing. The results here also provide current data on STBBI-related needs of GBM across cities beyond HIV. Testing prevalence in the past 6 months for STBBIs other than HIV for Engage is not markedly higher than previous estimates of testing for individual STBBIs (PHAC, 2011). Regarding STBBI testing, Montreal (74.9%) and Toronto GBM (67.4%) were less likely to report having been tested in the past 6 months for an STBBI other than HIV compared with Vancouver GBM (78.8%). The proportion of GBM with a history of syphilis infection ranged from 14% to 16% across cities. The consistency in syphilis antibody prevalence across sites was surprising given city differences in HIV diagnoses, and the high frequency of  syphilis/HIV coinfection (e.g., Gesink et al., 2014). Our findings are consistent with a widely recognized resurgence of syphilis among GBM from 2004 to 2015 in many Western countries (Abara et al., 2016). Our prevalence findings are consistent with other studies showing that GBM are disproportionately at risk for STBBI, especially HIV and syphilis (11.2 syphilis cases/100,000 in the Canadian general population; PHAC, 2020). The 13-18% of GBM who had never been tested for an STBBI other than HIV, combined with the syphilis history prevalence in our sample, suggests the need for increased efforts to reach GBM for STBBIrelated health services beyond HIV. Hepatitis testing among  GBM should be a continued priority, given that 3.9-7.3% have evidence of current or prior hepatitis C virus infection and that 0.3-1.3% have active hepatitis B virus infection in the current study. Regarding preventive and risk behaviours, PrEP use in the past 6 months among HIV-negative GBM was 9.6% in Montreal,11.4% in Toronto and 18.9% in Vancouver. It should be noted that the Montreal site completed data collection before the Toronto and Vancouver sites, which may have led to PrEP use being slightly lower in Montreal as PrEP use has continued to scale up in Canada following federal approval in 2016 (Mosley et al., 2018). PrEP use among HIVnegative GBM should be significantly higher in order to reduce HIV transmission as per clinical guidelines (Tan et al., 2017). However, across all cities, SDCAS in the past 6 months was relatively common (36.9-42.5% of GBM). Given the higher proportion of GBM in Vancouver engaging in SDCAS and greater PrEP use, focused efforts on testing and treating PrEP users for STBBI beyond HIV continue to be important.

Strengths and limitations
Our study represents GBM from the three geographic areas of highest HIV prevalence in Canada. Thanks to the RDS design (Heckathorn, 2002), the estimates presented here may be more accurate than in convenience samples due to our ability to correct for selection biases by adjusting estimates. Engage also extends previous research by combining self-report with biological measurement and the use of consistent measures across the three largest cities in Canada.
We should note that the data here present lifetime syphilis infection and lifetime hepatitis C infection as opposed to current infection. The use of throat and rectum culture testing for gonorrhea/chlamydia for half of the Toronto sample, and the missing data for 20% of rectal samples in Vancouver may have underestimated prevalence for these STIs in these two cities. Montreal's longer RDS recruitment chains may also have produced a more generalizable sample, including men who may have been less likely to have been reached in other studies. The differences in length of recruitment period between sites may have also led to biases in STBBI point prevalence estimates. We also note that while these data may approximate a probabilistic sample because estimates have been adjusted for RDS recruitment, the data are not adjusted for other factors such as sociodemographic or temporal differences (e.g., syphilis outbreaks), which could affect STBBI comparisons across cities.
Methodologically, our requirement for participants to complete STBBI testing may have underrepresented those who are not interested in getting tested. The smaller number of seed participants and larger sample size of the Montreal sample may have led to more reliable estimates in Montreal compared with Toronto or Vancouver, including the lower HIV prevalence found in Montreal. Given that provinces enacted HIV-related policies at different times and with different levels of access, it is unknown whether there is a need to focus programs differently across cities. Many of the outcomes assessed will therefore require further analyses and more detailed modelling, such as examining how HIV prevalence differences could be explained by differences in age or race across the three samples.

Conclusion
Engage is the first study in approximately a decade to present HIV and other STBBI prevalence using biological data in a representative sample of urban GBM in Canada. We identified some similarities in sexual health outcomes among GBM in the largest cities in Canada such as history of syphilis infection, but also notable differences, such as in HIV prevalence and PrEP uptake. It is essential to scale up both PrEP and STBBI testing in order to impact HIV and STBBI transmission among Canadian GBM. Acknowledgements The authors would like to thank the participants and staff of the Engage Study and our community engagement committees and community partner organizations.

Contributions to knowledge
Code availability Access to the data and code are available on request.
Author contributions TAH, GL, JC, DM, NT, DG and JJ originally conceptualized and drafted the present study. All of the authors made substantial contributions to the design of the study, revised the study critically for important intellectual content, gave approval for the version to be published, and agreed to be accountable for all aspects of the study herein. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.