Secondary Logo

Journal Logo


Belief in Treatment as Prevention and Its Relationship to HIV Status and Behavioral Risk

Card, Kiffer G. BSc*,†; Armstrong, Heather L. PhD*,‡; Lachowsky, Nathan J. PhD*,§,‖; Cui, Zishan MSc*; Sereda, Paul BA*; Carter, Allison MPH*,†; Montaner, Julio S. G. PhD*,‡; Hogg, Robert S. PhD*,†; Roth, Eric A. PhD‖,¶; Moore, David M. MD*,‡

Author Information
JAIDS Journal of Acquired Immune Deficiency Syndromes: January 1, 2018 - Volume 77 - Issue 1 - p 8-16
doi: 10.1097/QAI.0000000000001557



Highly active antiretroviral therapies (HAART) are enabling people living with HIV to manage their viral load and achieve viral suppression.1,2 HAART has thus contributed to significant declines in HIV-related morbidity, mortality, and transmission,3–6 giving rise to a prevention strategy known as “treatment as prevention” (TasP).7 Under TasP programs, the expansion of HAART to all persons living with HIV has been associated with lower rates of forward transmission, demonstrating the effectiveness of using biomedical prevention strategies to reduce HIV incidence.7–12

Yet, gay, bisexual, and other men who have sex with men (GBM) continue to be overrepresented in the HIV epidemic.13 In British Columbia (BC), an increasing proportion of new HIV cases occur among GBM, despite the number of new cases among GBM remaining relatively stable over the past decade.14 Given that biomedical prevention strategies, like TasP, have the potential to be highly effective,15 it is increasingly important that we understand their uptake and diffusion in at-risk social networks. Recent analyses of cross-sectional data collected between 2012 and 2014 demonstrated that although almost half of sampled GBM had heard of TasP, only 21% of HIV-positive men and 13% of HIV-negative men showed complete understanding of the concept.16 Other studies highlight a number of factors associated with TasP awareness among GBM, including HIV status, socioeconomic indicators, social embeddedness, engagement in health care, and substance use.16–18 However, as most studies on TasP have been cross-sectional, it remains uncertain which factors are associated with the longitudinal development of TasP-related awareness, attitudes, and behavior.

In addition to providing insight into the potential obstacles to GBM's knowledge of TasP, longitudinal research may also be relevant to understanding the emergence of pre-exposure prophylaxis (PrEP), which has likewise been shown to be a highly effective risk reduction measure limited by restricted access and uptake.19–22 Although we might not necessarily expect the diffusion of TasP and PrEP to be analogous,23 understanding factors that have shaped the diffusion of TasP may provide foresight for the eventual widespread rollout of PrEP.24 Indeed, based on Rogers diffusion of innovations theory,25 the spread of behavioral, ideological, and technological innovations may share common patterns of diffusion. For instance, one study reported that doctors who defer treatment for newly HIV-positive patients are also skeptical about PrEP.24 This demonstrates potential similarities in the diffusion of these biomedical prevention strategies, highlighting the need to examine the diffusion of prevention attitudes within GBM's social networks.


Study Procedures

Since 1992, the distribution of HIV medications has been managed centrally by the British Columbia Centre for Excellence in HIV/AIDS and thereby made freely available to all persons living with HIV, regardless of CD4 count or viral load.26 In 2010, BC initiated a pilot program, known as STOP HIV/AIDS, to expand HIV testing and facilitate engagement in care and treatment for people living with HIV. The Momentum Health Study was established to investigate the effects of this expansion. Accordingly, data for these analyses were collected between 2012 and 2016. Before recruitment, formative research was undertaken to evaluate the social network structure of GBM in Metro Vancouver.27 Based on community maps, 134 initial participants, referred to as seeds, were recruited with the help of community partners and by advertising on popular geosocial networking applications tailored for GBM.28 Seeds were then given 6 vouchers and asked to recruit other eligible participants (ie, sexually active GBM, aged ≥16 years, living in Metro Vancouver, BC). Respondent-driven sampling (RDS) continued with each new recruit receiving up to 6 vouchers that they could provide to other GBM in their networks. Participants provided informed consent and were invited to enroll in a cohort with follow-up at 6-month intervals.

At each study visit, participants completed a computer-assisted self-interview, which assessed demographic, biobehavioral, and psychosocial factors. Upon completion of the questionnaire, participants underwent clinical screenings during which a study nurse administered a point-of-care HIV test or CD4 and viral load blood draws as appropriate, venipuncture for hepatitis C and syphilis testing, and optional screenings for gonorrhea and chlamydia. Upon completion of each study visit, participants received a $50 CAD honoraria, which could be taken as cash or used to purchase $10 tickets for monthly and semiannual drawings. Ethics approval was received from the research ethics boards at the Simon Fraser University, the University of British Columbia, and the University of Victoria.

Statistical Analysis

All statistical analyses were conducted in SAS version 9.4. (SAS Institute Inc., Cary, NC) TasP knowledge, attitudes, and behavior were characterized using repeated-measures latent class analysis (RMLCA). Unconditional RMLCA models were constructed using the PROC LCA routine developed by Lanza and Collins.29 Considering the difference in treatment and HIV-awareness between HIV-negative/unknown and HIV-positive groups,16 all analyses were stratified by self-reported HIV status. This was statistically supported because the assumption of measurement invariance restriction was rejected during RMLCA model specification. In specifying the number of classes in the RMLCA model, the adjusted Bayesian information criterion fit statistic was used, as it has been shown to perform well in LCA modeling.30 Ultimately, however, model selection was made with consideration of theoretical and conceptual appropriateness. Supplemental Digital Content Figure 1, provides a chart showing statistical fit indices for models constructed with 2, 3, 4, and 5 classes. Accordingly, a 3-class RMLCA solution was selected because it minimized the Akaike information criterion, Bayesian information criterion, and adjusted Bayesian information criterion for HIV-negative men and provided a class structure that was interpretable, parsimonious, and conceptually appropriate for both groups.

Bivariate odds ratios and 95% confidence intervals were generated using generalized equation modeling to identify the variables associated with class membership. Class membership was defined by posterior class membership probabilities, generated by the PROC LCA routine. Variables with a bivariate P value of <0.20 were included in a multivariable model for each class. Although multinomial regression models are more common in LCA, we constructed separate binary logistic models using generalized estimating equations for each class-by-class comparison; given a 3-class solution, this meant building 3 models, each with the same reference class. This modeling approach allowed the predictors of class membership to load independently for each class, providing independent and adjusted associations for each latent class rather than for the broad general construct. Each of these final models were constructed using a backwards elimination procedure to identify and adjust the independent covariates associated with TasP endorsement (RMLCA class membership). At each step, the variable with the lowest type-III P value was removed from the model until an optimal (minimized) Quasi-Akaike information criterion value was achieved. As this study was longitudinal, additional bivariate analyses were undertaken to test for an interaction between TasP endorsement (LCA class membership) and visit number (time), allowing us to identify the factors associated with “slower” and “faster” uptake of TasP-related attitudes, knowledge, and behavior over time.


Explanatory variables included participant demographics, sexual behavior, substance use, social behavior, and clinical outcomes. Demographic variables included age group, sexual orientation, race/ethnicity, education, annual income, and relationship status. Sexual behavior variables included recent (past 6 months, P6M) condomless anal sex (CAS), number of anal sex partners, and recent sexually transmitted infections (STI) diagnosis, as well as whether participants reported “always using condoms.” Substance use variables included P6M use of cannabis, inhalants, stimulants, sedatives, opioids, poppers, erectile dysfunction drugs, party drugs [ie, methamphetamines, ecstasy, ketamine, mushrooms, lysergic acid diethylamide (LSD), other hallucinogens, gamma-hydroxybutyric acid], injection drugs (excluding steroids), and number of binge drinking events in the past month (defined by having more than 5 drinks on a single occasion). Social behavior variables included the amount of social time participants spent with other GBM and the number of GBM they reported knowing. Clinical variables included past year pharmacy refill adherence data, CD4 counts, and HIV viral load (<200, ≥200 copies/mL).


In total, 556 HIV-negative/unknown men provided 1845 observations and 218 HIV-positive men provided 745 observations. Among participants with at least 1 follow-up visit (n = 575), the median follow-up time was 1.98 years (Q1–Q2: 1.49–2.49 years). At baseline, the median age of the sample was 34 years (Q1–Q3: 26–47 years) and the majority of the sample identified as gay (84.6%), were white (75.6%), had at least some postsecondary education (76.9%), and had annual incomes of <Can $30,000 (62.7%). Over the P6M, 65.3% (n = 504/772) reported having CAS, the median number of anal sex partners was 3 (Q1–Q3: 1–8), and 10.6% reported being diagnosed with an STI. Of those reporting CAS, 60.9% (n = 307/504) reported CAS with at least 1 seroconcordant/unknown partner, 50.0% (n = 252/504) reported CAS with at least 1 unknown status partner, and 22.4% (n = 113/504) reported CAS with at least 1 serodiscordant partner. Table 1 provides additional descriptive statistics for the sample, stratified by self-reported HIV status.

Person-Level Descriptive Results for the Momentum Study

As described earlier, a 3-class RMLCA model was used to characterize the multidimensional patterns of TasP endorsement (Table 2). In brief, the 3 resultant classes can be characterized as (1) those unaware of TasP, (2) those who were skeptical of TasP, and (3) those who were believing of TasP efficacy.

Three-Class Latent Class Solutions Showing the Distribution of Repeated Measure Indicator Responses, Stratified by Self-Reported HIV-Status

Class 1: Unaware

The majority (64.2%) of HIV-negative/unknown men and one-third (29.2%) of HIV-positive men belonged to the “unaware” latent class. This class was characterized as not having heard of TasP, not engaging in viral load sorting, and being generally skeptical of the preventive benefits of TasP. For example, only 10.3% of HIV-negative/unknown men and 7.8% of HIV-positive men in this class agreed that “a person with an undetectable viral load cannot pass on the virus.”

Class 2: Skeptical

The second latent class, described as the “Skeptical” class, represented 29.7% of HIV-negative/unknown men and 23.1% of HIV-positive men. The Skeptical class was characterized by greater awareness of TasP compared with the Unaware class but also greater skepticism compared with the Believing class. For example, only 10.3% HIV-negative/unknown men and 53.9% of HIV-positive men in this class engaged in viral load sorting; 91.8% of Skeptical HIV-positive men also did not believe that an undetectable viral load was associated with an inability to transmit the virus; 37.3% of HIV-positive men and 37.0% HIV-negative/unknown men felt undetectable men still had to worry about transmitting HIV to their partners, and a significant proportion of both HIV-positive (31.9%) and HIV-negative/unknown (41.6%) men in this class felt that CAS was never safe regardless of viral load.

Class 3: Believing

The final latent class represented 47.7% of HIV-positive men but only 6.1% of HIV-negative/unknown men. Men in the Believing class reported that TasP lowered HIV risk “a lot,” engaged in condomless viral load sorting with their serodiscordant partners, and agreed with statements regarding the preventive benefits of TasP. Despite this, 58% men in the Believing class believed that condomless sex is “never safe” regardless of viral load.

The multivariable models for HIV-positive and HIV-negative/unknown men are provided in Table 3. Compared with Unaware HIV-negative/unknown men, Skeptical HIV-negative/unknown men were more likely to identify their sexual orientation as “other,” to be white to report recent erectile dysfunction and party drug use, and to spend more social time with other gay men; they were also less likely to report “always using condoms.” Skeptical HIV-positive men were more likely than Unaware HIV-positive men to be aged 30–39 years (vs. 16–29 years), to have CAS, and to spend more than 3 quarters of their social time with other gay men (vs. ≤25%); they were also less likely to always use condoms. Again, compared with the Unaware class, Believing HIV-negative/unknown men were more likely to identify as gay (vs. bisexual), engage in CAS, report recent erectile dysfunction and party drug use, and spend more than 3 quarters (vs. ≤25%) of their social time with other GBM; they were less likely to report “always using condoms.” Believing HIV-positive men were more likely than Unaware HIV-positive men to engage in serodiscordant or unknown CAS and to have more anal sex partners; they were less likely to binge drink.

Multivariable Models Showing the Independent and Adjusted Correlates of Class Membership

Over time, TasP endorsement was relatively stable among HIV-positive men (P = 0.20). However, HIV-negative/unknown men became increasingly aware of TasP (P < 0.0001). Figure 1 shows the proportion of men belonging to each class at each study visit. In summary, the proportion of Unaware HIV-negative men dropped by 40.1%, whereas the Skeptical class increased 29.1% and the Believing class increased 11.0%. For HIV-positive men, the proportion in the Unaware class dropped 4.3%, the proportion in the Skeptical class dropped 4.6%, and the proportion in Believing class increased 8.9%.

RMLCA class membership, by visit number.

Using bivariate interaction terms (Supplemental Digital Content Table 2,, we examined the relationship between TasP endorsement and time. Among HIV-negative/unknown men, we determined that having greater than a high school education [odds ratio (OR): 1.16; 95% confidence interval (CI): 1.01 to 1.34], being employed (OR: 1.13; 95% CI: 1.01 to 1.26), and having a regular partner (OR: 1.14; 95% CI: 1.03 to 1.26) were associated with a respective 16%, 13%, and 14% increase in odds of transitioning to a more-aware class, per visit (ie, faster TasP endorsement). Conversely, among HIV-negative/unknown men, reporting recent CAS (OR: 0.86; 95% CI: 0.76 to 0.98) or STI diagnosis (OR: 0.67; 95% CI: 0.50 to 0.89) was associated with a respective 14% and 33% decrease in odds of transitioning to a more-aware class, per visit (ie, slower TasP endorsement). For HIV-positive men, being antiretroviral therapy naive or recently starting treatment in the past 12 months (OR: 1.45; 95% CI: 1.09 to 1.94), being older (≥40 vs. 16–29; OR: 2.13; 95% CI: 1.13 to 3.99), and always using condoms (OR: 1.24; 95% CI: 1.09 to 1.40) were associated with a respective 113% and 24% increase in odds of transitioning to a more-aware class, per visit (ie, faster TasP endorsement). Being Asian (vs. white; OR: 0.72; 95% CI: 0.57 to 0.90), use of inhalants (OR: 0.49; 95% CI: 0.45 to 0.53), sedatives (OR: 0.81; 95% CI: 0.68 to 0.97), and poppers (OR: 0.86; 95% CI: 0.76 to 0.97) was associated with a respective 28%, 51%, 19%, and 14% decrease in odds of transitioning to a more-aware class, per visit (ie, slower TasP endorsement).


Using RMLCA, we identified 3 classes representing a continuum of increasing TasP endorsement among 774 GBM recruited using RDS in Metro Vancouver: (1) those unaware of TasP, (2) those who were aware though skeptical of TasP, and (3) those who were believing of TasP efficacy. Interpreting our results within a diffusion of innovations framework,25 our study highlights that although some TasP concepts, such as the importance of knowing an HIV-positive partner's viral load, have diffused within the observed sociosexual networks, other concepts such as the preventive benefits of TasP have had a more limited impact. This is consistent with the past research that has found that polarizing biomedical prevention strategies can have limited diffusion.18

Consistent with previous research on the longitudinal trends of TasP awareness,31 our study found that TasP support was consistently higher among HIV-positive men than HIV-negative men throughout the study period. Accordingly, the largest class transitions were seen among HIV-negative men. There was a 40% decrease in the proportion of Unaware HIV-negative men, compared with a 4% decrease in Unaware HIV-positive men. However, most HIV-negative men transitioned from the Unaware class to the Skeptical class, and among both HIV-negative and HIV-positive men, the proportion of those transitioning into the Believing classes was only around 10%. These results can be considered in light of previous research, which has reported skepticism among HIV-negative men about whether an undetectable viral load can prevent HIV transmission.32 However, HIV-positive men in the Believing class were actually slightly less optimistic of TasP's ability to eliminate transmission worries compared with HIV-negative men in the same class. Furthermore, we observed that HIV-positive men who recently started treatment or who had never been on treatment increased in TasP support faster than those with suboptimal adherence. This may suggest that HIV-positive men place greater weight on the potential limitations of TasP—perhaps owing to their greater exposure to medical professionals, sexual health information, and social norms that place a disproportionate responsibility on people living with HIV to prevent HIV transmission.

Our findings also present a nuanced view of the effect of TasP on sexual behaviors. We observed that greater TasP endorsement was associated with greater odds for engaging in any CAS and with the use of erectile dysfunction and party drugs. Furthermore, membership in classes representing higher TasP endorsement was associated with lower odds of reporting “always using condoms.” These 2 key findings suggest that TasP endorsement is associated with increased CAS—and thus, in the absence of PrEP or TasP,6,20 greater risk for HIV. This finding generally agrees with previous studies examining the impact of TasP on CAS and subsequent acquisition of STI.33–35 However, these findings should also be considered in light of research, indicating that men willing to rely on TasP and other biomedical prevention strategies are already engaging in CAS.36 Indeed, TasP endorsement may not necessarily be the most salient predictor of condom use.37 It may be that TasP merely provides GBM with the knowledge and comfort to rely on other risk management options,38 such as viral-load sorting and PrEP—which a growing body of research suggests may eliminate the risk for HIV during CAS.6,20

Our findings support the view that TasP awareness may actually be an indicator of greater health literacy. For instance, we observed that faster TasP endorsement was associated with decreased odds for CAS and having a recent STI diagnosis among HIV-negative men. Likewise, “always using condoms” was associated with faster, not slower, TasP endorsement. These findings suggest that TasP has diffused more effectively among men who had been using condoms and effectively preventing STIs but now employ TasP-related strategies as part of their risk management approach. With consideration to diffusions of innovation theory, this suggests that condoms may be increasingly viewed as a dated HIV prevention strategy by GBM who have adopted newer, preferred alternative risk management strategies, such as viral-load sorting and PrEP. With that said, we note that only about half of HIV-positive men in the Believing and Skeptical classes reported using their viral load status in determining when to use condoms with HIV-negative partners. Furthermore, a higher proportion of HIV-positive and HIV-negative men in the Believing classes agreed that condomless sex is “never safe” despite believing that people with undetectable viral loads “cannot pass on the virus” and “do not need to worry so much about infecting others with HIV.” These findings suggest that individuals well informed of TasP, and especially HIV-positive persons, continue to weigh their sexual health and pleasure-seeking behaviors carefully. Furthermore, their agreement that CAS is “never safe” may indicate that they recognize the limitations of biomedical HIV preventions in addressing the risk for other STIs. However, given the independent diffusion of TasP principles, these findings provide an example of the potential benefits of educating GBM of both the benefits and limitations of TasP, allowing them to leverage evidence-driven knowledge to better manage their sexual health.


Considering that the diffusion of TasP differed across key subgroups of GBM, which are regularly described as at risk (eg, socially isolated men, ethnic minorities, those with less formal education, those who are unemployed, and those who use sex and party drugs), ongoing and future biomedical interventions should be careful not to assume that the diffusion of prevention strategies will reach these communities. This is especially important given research that has found that differences in the sexual scripts and norms between subgroups can contribute to the breakdown of sexual negotiations, which otherwise provide GBM with opportunities to protect themselves and their partners from unwanted risks.39 Programs designed to ensure that all GBM have a basic understanding of the benefits and limitations of biomedical prevention might also have the added benefit of addressing lagging negative attitudes toward HIV-positive men, thereby helping to establish realistic risk perceptions for those who experience significant anxiety regarding their sexual behavior. Indeed, because disclosure of serostatus is an important precursor of several risk reductions strategies employed by GBM,40–42 reducing stigma for HIV may enhance disclosure self-efficacy among HIV-positive men, making these strategies more effective for all GBM.43,44 This is especially important for contexts where risk reduction measures not dependent on serodisclosure, such as PrEP, are not yet widely available. Furthermore, previous studies have also suggested that reduced stigma may facilitate better uptake of care—resulting ultimately in a higher proportion of HIV-positive men achieving undetectability.45,46 Among HIV-negative men, reduced HIV-stigma, brought about by increased TasP awareness, also has the potential to improve HIV testing by reducing the fear associated with an HIV-positive result.47 Research has also shown that increased TasP awareness is associated with taking postexposure prophylaxis—reinforcing the idea that greater treatment awareness can contribute to risk-reduction strategies among GBM who are already engaging in high-risk sexual behaviors.31,36


There are several important limitations of this study. First, because of the nature of our questions, these data are vulnerable to favorable reporting. Although computer-administered questionnaires may reduce some social pressures for favorable reporting, participants may be aware of the research implications of our work and therefore respond accordingly. Alternatively, some participants may not have fully recognized “treatment as prevention” as a descriptor for concepts they were in fact aware of—artificially reducing their endorsement of TasP. Second, by dichotomizing Likert-agreement scales for inclusion in the LCA, we may have masked over smaller transitions characterizing increasing TasP endorsement. Third, as the data are longitudinal, it is possible that increasing TasP awareness is attributable to repeated measures. However, considering that faster TasP endorsement occurred primarily among HIV-negative men, repeated measures is not the best explanation for the observed changes. Fourth, as the data were distributed by visit number, rather than across time, our longitudinal estimates are subject to error. Finally, as we did not adjust for RDS structure in our analysis, the independence of observations cannot be assumed, and it is possible that the observed associations are biased along some unmeasured network characteristic inherent to the RDS sampling methodology we have employed. With that said, we found that equilibrium was achieved across several of the key indicators monitored during RDS recruitment (eg, age, sexual orientation, ethnicity, citizenship, residence, student status, and income)—suggesting that with respect to these indicators, our sample was not biased based on our initial recruits and was in fact likely representative of the social networks being studied in the present analysis.48 We should note, however, that HIV-positive men were deliberately oversampled. However, because our sample is stratified by HIV status, we would not expect this to impact our study results.


Despite these limitations, the present study highlights increasing TasP awareness among Vancouver GBM with significant increases among HIV-negative/unknown GBM. However, individuals in several key groups (eg, substance users, those with low socioeconomic status, and those with lower attachment to other GBM) remain unaware of TasP and may therefore misunderstand how to safely negotiate sexual behavior in TasP environments. We conclude that it is likely that the diffusion of other HIV prevention strategies, such as PrEP, may likewise be hindered among these groups. Therefore, public health programs should seek to overcome the overall lack of health literacy among GBM with the aim of establishing broad, compatible safe-sex norms throughout at-risk social networks.


The authors thank the Momentum Health Study participants, office staff and community advisory board, as well as our community partner agencies, Health Initiative for Men, YouthCO HIV & Hep C Society, and Positive Living Society of BC.


1. Graham SM, Holte SE, Peshu NM, et al. Initiation of antiretroviral therapy leads to a rapid decline in cervical and vaginal HIV-1 shedding. AIDS. 2007;21:501–507.
2. Vernazza PL, Troiani L, Flepp MJ, et al. Potent antiretroviral treatment of HIV-infection results in suppression of the seminal shedding of HIV. The Swiss HIV Cohort Study. AIDS. 2000;14:117–121.
3. Antiretroviral Therapy Cohort. Life expectancy of individuals on combination antiretroviral therapy in high-income countries: a collaborative analysis of 14 cohort studies. Lancet. 2008;372:293–299.
4. Palella FJ, Delaney KM, Moorman AC, et al. Declining morbidity and mortality among patients with advanced human immunodeficiency virus infection. HIV Outpatient Study Investigators. N Engl J Med. 1998;338:853–860.
5. Cohen MS, Chen YQ, McCauley M, et al. Prevention of HIV-1 infection with early antiretroviral therapy. N Engl J Med. 2011;365:493–505.
6. Rodger A, Cambiano V, Brunn T, et al. Sexual activity without condoms and risk of HIV transmission in serodifferent couples when the HIV-positive partner is using suppressive antiretroviral therapy. JAMA. 2016;316:171–181.
7. Montaner JSG, Lima VD, Harrigan PR, et al. Expansion of HAART coverage is associated with sustained decreases in HIV/AIDS morbidity, mortality and HIV transmission: the “HIV treatment as prevention” experience in a Canadian setting. PLoS One. 2014;9:e87872.
8. Das M, Chu PL, Santos GM, et al. Decreases in community viral load are accompanied by reductions in new HIV infections in San Francisco. PLoS One. 2010;5:e11068.
9. Lima VD, Lourenço L, Yip B, et al. Trends in AIDS incidence and AIDS-related mortality in British Columbia between 1981 and 2013. Lancet HIV. 2015;2:e92–e97.
10. Lima VD, Eyawo O, Ma H, et al. The impact of scaling-up combination antiretroviral therapy on patterns of mortality among HIV-positive persons in British Columbia, Canada. J Int AIDS Soc. 2015;18:20261.
11. Montaner JSG, Lima VD, Barrios R, et al. Association of highly active antiretroviral therapy coverage, population viral load, and yearly new HIV diagnoses in British Columbia, Canada: a population-based study. Lancet. 2010;376:532–539.
12. Nosyk B, Min J, Lima V, et al. Modelling the cost-effectiveness of population-level HAART expansion in British Columbia. Lancet HIV. 2015;2:e393–e400.
13. Public Health Agency of Canada. Population-Specific HIV/AIDS Status Report: Gay, Bisexual, Two-Spirit and Other Men Who Have Sex With Men. Ottawa, ON, Canada: Public Health Agency of Canada; 2014. Available at: Accessed August 4, 2015.
14. British Columbia Provincial Health Officer. HIV, Stigma and Society: Tackling a Complex Epidemic and Renewing HIV Prevention for Gay and Bisexual Men in British Columbia. Victoria, BC, Canada: British Columbia Provincial Health Officer; 2014. Available at: Accessed August 4, 2015.
15. UNAIDS. Fast-Tracking Combination Prevention: Towards Reducing New HIV Infections to Fewer than 500,000 by 2020. Geneva, Switzerland: UNAIDS; 2015. Available at: Accessed August 4, 2015.
16. Carter A, Lachowsky NJ, Rich A, et al. Gay and bisexual men's awareness and knowledge of treatment as prevention: findings from the Momentum Health Study in Vancouver, Canada. J Int AIDS Soc. 2015;18:20039.
17. Prati G, Zani B, Pietrantoni L, et al. PEP and TasP awareness among Italian MSM, PLWHA, and high-risk heterosexuals and demographic, behavioral, and social correlates. PLoS One. 2016;11:e0157339.
18. Nodin DN, Carballo-Diéguez A, Ventuneac AM, et al. Knowledge and acceptability of alternative HIV prevention bio-medical products among MSM who bareback. AIDS Care. 2008;20:106–115.
19. Burns DN, Grossman C, Turpin J, et al. Role of oral pre-exposure prophylaxis (PrEP) in current and future HIV prevention strategies. Curr HIV/AIDS Rep. 2014;11:393–403.
20. Grant RM, Lama JR, Anderson PL, et al. Preexposure chemoprophylaxis for HIV prevention in men who have sex with men. N Engl J Med. 2010;363:2587–2599.
21. Golub SA, Gamarel KE, Rendina HJ, et al. From efficacy to effectiveness: facilitators and barriers to PrEP acceptability and motivations for adherence among MSM and transgender women in New York city. AIDS Patient Care STDs. 2013;27:248–254.
22. Bauermeister JA, Meanley S, Pingel E, et al. PrEP awareness and perceived barriers among single young men who have sex with men. Curr HIV Res. 2013;11:520–527.
23. Krakower D, Ware N, Mitty JA, et al. HIV providers' perceived barriers and facilitators to implementing pre-exposure prophylaxis in care settings: a qualitative study. AIDS Behav. 2014;18:1712–1721.
24. Krakower DS, Beekmann SE, Polgreen PM, et al. Diffusion of newer HIV prevention innovations: variable practices of frontline infectious diseases physicians. Clin Infect Dis Off Publ Infect Dis Soc Am. 2016;62:99–105.
25. Rogers EM. Diffusion of Innovations, 5th ed. New York, NY: Simon and Schuster; 2003.
26. Gustafson R, Montaner J, Sibbald B. Seek and treat to optimize HIV and AIDS prevention. Can Med Assoc J. 2012;184:1971.
27. Forrest JI, Stevenson B, Rich A, et al. Community mapping and respondent-driven sampling of gay and bisexual men's communities in Vancouver, Canada. Cult Health Sex. 2014;16:288–301.
28. Lachowsky NJ, Lal A, Forrest JI, et al. Including online-recruited seeds: a respondent-driven sample of men who have sex with men. J Med Internet Res. 2016;18:e51.
29. Lanza ST, Collins LM. A mixture model of discontinuous development in heavy drinking from ages 18 to 30: the role of college enrollment. J Stud Alcohol. 2006;67:552–561.
30. Nylund KL, Asparouhov T, Muthén BO. Deciding on the number of classes in latent class analysis and growth mixture modeling: a Monte Carlo Simulation Study. Struct Equ Model Multidiscip J. 2007;14:535–569.
31. Holt M, Lea T, Schmidt H-M, et al. Increasing belief in the effectiveness of HIV treatment as prevention: results of repeated, National Surveys of Australian Gay and Bisexual Men, 2013–15. AIDS Behav. 2016;20:1564–1571.
32. Holt M, Murphy D, Callander D, et al. HIV-negative and HIV-positive gay men's attitudes to medicines, HIV treatments and antiretroviral-based prevention. AIDS Behav. 2013;17:2156–2161.
33. Delva W, Helleringer S. Beyond risk compensation: clusters of antiretroviral treatment (ART) users in sexual networks can modify the impact of ART on HIV incidence. PLoS One. 2016;11:e0163159.
34. Eaton LA, Kalichman SC. Risk compensation in HIV prevention: implications for vaccines, microbicides, and other biomedical HIV prevention technologies. Curr HIV/AIDS Rep. 2007;4:165–172.
35. Katz MH, Schwarcz SK, Kellogg TA, et al. Impact of highly active antiretroviral treatment on HIV seroincidence among men who have sex with men: San Francisco. Am J Public Health. 2002;92:388–394.
36. Bavinton BR, Holt M, Grulich AE, et al. Willingness to act upon beliefs about “treatment as prevention” among Australian gay and bisexual men. PLoS One. 2016;11:e0145847.
37. Lachowsky NJ, Tanner Z, Cui Z, et al. An event-level analysis of condom use during anal intercourse among self-reported human immunodeficiency virus-negative gay and bisexual men in a treatment as prevention environment. Sex Transm Dis. 2016;43:765–770.
38. Otis J, McFadyen A, Haig T, et al. Beyond condoms: risk reduction strategies among gay, bisexual, and other men who have sex with men receiving rapid HIV testing in Montreal, Canada. AIDS Behav. 2016;20:2812–2826.
39. Adam BD, Husbands W, Murray J, et al. Silence, assent and HIV risk. Cult Health Sex. 2008;10:759–772.
40. Card KG, Lachowsky NJ, Cui Z, et al. Seroadaptive strategies of gay & bisexual men (GBM) with the highest quartile number of sexual partners in Vancouver, Canada. AIDS Behav. 2016;21:1452–1466.
41. Card KG, Lachowsky NJ, Cui Z, et al. A Latent class analysis of seroadaptation among gay and bisexual men. Arch Sex Behav. 2016:1–12. doi: 10.1007/s10508-016-0879-z.
42. Persson A. Sero-silence and sero-sharing: managing HIV in serodiscordant heterosexual relationships. AIDS Care. 2008;20:503–506.
43. Skinta MD, Brandrett BD, Schenk WC, et al. Shame, self-acceptance and disclosure in the lives of gay men living with HIV: an interpretative phenomenological analysis approach. Psychol Health. 2014;29:583–597.
44. Evangeli M, Wroe AL. HIV disclosure anxiety: a systematic review and theoretical synthesis. AIDS Behav. 2017;21:1–11.
45. King R, Katuntu D, Lifshay J, et al. Processes and outcomes of HIV serostatus disclosure to sexual partners among people living with HIV in Uganda. AIDS Behav. 2008;12:232–243.
46. Ekama SO, Herbertson EC, Addeh EJ, et al. Pattern and determinants of antiretroviral drug adherence among Nigerian pregnant women. J Pregnancy. 2012;2012:851810.
47. Knussen C, Flowers P, McDaid LM. Factors associated with recency of HIV testing amongst men residing in Scotland who have sex with men. AIDS Care. 2014;26:297–303.
48. Heckathorn DD. Snowball versus respondent driven sampling. Sociol Methodol. 2011;41:355–366.

HIV; gay and bisexual men; sexual behavior; HAART; harm reduction

Supplemental Digital Content

Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved.