UBC Faculty Research and Publications

High HIV risk and syndemic burden regardless of referral source among MSM screening for a PrEP demonstration… Wilton, James; Noor, Syed W; Schnubb, Alexandre; Lawless, James; Hart, Trevor A; Grennan, Troy; Fowler, Shawn; Maxwell, John; Tan, Darrell H S Feb 27, 2018

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


52383-12889_2018_Article_5180.pdf [ 913.37kB ]
JSON: 52383-1.0364059.json
JSON-LD: 52383-1.0364059-ld.json
RDF/XML (Pretty): 52383-1.0364059-rdf.xml
RDF/JSON: 52383-1.0364059-rdf.json
Turtle: 52383-1.0364059-turtle.txt
N-Triples: 52383-1.0364059-rdf-ntriples.txt
Original Record: 52383-1.0364059-source.json
Full Text

Full Text

RESEARCH ARTICLE Open AccessHigh HIV risk and syndemic burdenregardless of referral source among MSMscreening for a PrEP demonstration projectin Toronto, CanadaJames Wilton1, Syed W. Noor2, Alexandre Schnubb3, James Lawless3, Trevor A. Hart2,4, Troy Grennan5,6,Shawn Fowler7, John Maxwell8 and Darrell H. S. Tan3,9*AbstractBackground: To maximize public health impact and cost-effectiveness, HIV pre-exposure prophylaxis (PrEP) mustreach individuals at high HIV risk. Referrals for PrEP can be self- or provider-initiated, but there are several challengesto both. We assessed whether HIV risk differed by referral source among gay, bisexual and other men who have sex(gbMSM) screening for an HIV PrEP demonstration project.Methods: PREPARATORY-5 was an open-label PrEP demonstration project enrolling gbMSM at high risk of HIVacquisition in Toronto, Canada. Study eligibility criteria related to high risk was defined as scoring ≥10 on the HIVIncidence Risk Index for MSM (HIRI-MSM) and engaging in at least 1 act of condomless receptive anal sex withinthe past 6 months. Recruitment was promoted through self-referrals (ads in a sexual networking app and gaynewspaper/website) and provider-referrals (10 community-based organizations, CBOs). HIV risk score (HIRI-MSM) andsyndemic health burden were measured among gbMSM screened for study participation and compared accordingto referral source.Results: Between October 16 and December 30, 2014, online ads generated 1518 click-throughs and CBOs referred115 individuals. Overall, 165 men inquired about the trial, of which 86 underwent screening. The majority ofscreened men were self-referrals (60.5%), scored ≥10 on HIRI-MSM (96.5%), and reported condomless receptive analsex in the past 6 months (74.2%). Self- and provider-referrals had similarly high HIV risk profiles, with a median (IQR)HIRI-MSM score of 26.0 (19.0–32.5) and 28.5 (20.0–34.0) (p = 0.3), and 75.0% and 73.5% reporting condomlessreceptive anal sex (p = 0.9), respectively. The overall burden of syndemic health problems was also high, withapproximately one-third overall identified as having depressive symptoms (39.5%), alcohol-related problems (39.5%),multiple drug use (31.4%), or sexual compulsivity (31.4%). There were no significant differences in syndemic healthproblems by referral source.Conclusions: HIV risk and syndemic burden were high among gbMSM presenting for this PrEP demonstrationproject regardless of referral source. Self-referral may be a useful and efficient strategy for identifying individualssuitable for PrEP use. Online strategies and CBOs working in gay men’s health may play important roles inconnecting individuals at high HIV risk to PrEP services.Trial registration: ClinicalTrials.gov NCT02149888. Registered May 12th 2014.Keywords: PrEP, Gay men and other men who have sex with men, Syndemics, HIV risk, Screening* Correspondence: darrell.tan@gmail.com3Division of Infectious Diseases, St. Michael’s Hospital, 30 Bond St, 4CC –Room 4-179, Toronto, ON M5B 1W8, Canada9Department of Medicine, University of Toronto, Toronto, Ontario, CanadaFull list of author information is available at the end of the article© The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, andreproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link tothe Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.Wilton et al. BMC Public Health  (2018) 18:292 https://doi.org/10.1186/s12889-018-5180-8BackgroundDaily oral tenofovir/emtricitabine as pre-exposure prophy-laxis (PrEP) is highly effective at preventing HIV amongmen who have sex with men when used as directed [1, 2],and was approved by the U.S. Food and Drug Administra-tion in 2012 and Health Canada in 2016. Guidelines fromthe World Health Organization [3], U.S. Centers forDisease Control and Prevention [4], and others [5],recommend restricting PrEP to those at high risk of HIVinfection in order to maximize public health benefits andcost-effectiveness. As implementation proceeds, a majorchallenge lies in identifying such individuals and referringthem to PrEP services.PrEP referrals can be either self- or provider- initiated,but each strategy has potential challenges. Providers mayhave trouble recognizing or assessing a client’s risk ofHIV infection, in part because many HIV risk behavioursare stigmatized and therefore not discussed [6], leadingto missed opportunities for referral. Indeed, many gay,bisexual and other men who have sex with men(gbMSM) do not feel comfortable discussing their sexualhealth with providers [7, 8], and even when HIV risk isdiscussed it can be difficult to assess due to the widerange of biological, behavioural, and social factors in-volved. Clinical HIV risk scoring tools, such as the HIVIncidence Risk Index for Men who have Sex with Men(HIRI-MSM) [9], have been developed to facilitate dis-cussions about HIV risk behaviours and assist providersin identifying those at highest risk, but it is unclear towhat extent these tools are used.On the other hand, at-risk individuals may seek outPrEP themselves (self-refer), but we and others [7, 10, 11]have previously identified low perceived HIV risk amongobjectively high risk gay men (as determined by HIRI-MSM score [7], having undiagnosed HIV infection [10],or meeting the entry criteria for the iPrEX PrEP clinicaltrial [11]) as a potential barrier to such referrals. Possiblereasons for the discordance between objective and per-ceived HIV risk include lack of knowledge, health-relatedoptimism, and denial/avoidance (due to fear and shamecreated by stigma/discrimination) [12, 13]. Conversely,some gbMSM who self-refer may be at low risk forHIV but highly anxious about HIV transmission, orbe individuals who want to use PrEP as an alternativeto their pre-existing consistent condom use. PrEP usein such individuals may sometimes be clinically justi-fiable, but widespread uptake in these groups couldundermine the cost-effectiveness of PrEP [14]. This isbecause more of such individuals would need to beusing PrEP in order to avert a single HIV infection(ie. higher number needed to treat), thus incurringgreater financial costs per infection prevented.To explore the effectiveness of different referral strat-egies in directing high risk gbMSM to PrEP in Toronto,Canada, we compared the level of HIV risk amonggbMSM screened for a pilot PrEP demonstration projectaccording to whether they were self- or provider-referred. To measure HIV risk as comprehensively aspossible, we used both a sexual behaviour-based index ofHIV risk in gbMSM (the HIRI-MSM screening tool) andvalidated scales that screen for mental health problems.We compared the burden of co-occurring mental healthproblems in these groups because problems such as de-pression and substance use are tightly associated withHIV risk among gbMSM [15, 16], and because linkingPrEP patients to care for such “syndemic” problems mayreduce underlying HIV risk and facilitate PrEP adher-ence. Further, the HIRI-MSM only includes items onsexual and drug-taking behaviors, and does not includeitems for mental health problems and other more distalHIV risk factors, despite their objective connection toHIV risk. We have previously observed a high preva-lence of syndemic problems among gay men seekingboth PrEP [17] and nPEP (non-occupational post-exposure prophylaxis) [18], but these analyses were notstratified by referral type. In this study, we hypothesizedthat provider-referred men would have a higher HIV riskprofile (as determined by their HIRI-MSM score andnumber of syndemic problems) than men who self-referred.MethodsTrial and eligibility criteriaPREPARATORY-5 was an open-label, 12-month demon-stration project at an academic hospital-based HIV clinicin downtown Toronto whose main objectives were toobtain pilot data on PrEP acceptability and clinical out-comes among Toronto gbMSM (NCT02149888). To beeligible, gbMSM had to be 18 years or older, living inthe greater Toronto area, HIV-negative, and at high riskfor HIV acquisition (defined as scoring ≥10 on the HIRI-MSM [9] and engaging in at least 1 act of condomlessreceptive anal sex within the past 6 months). The targetsample size for the pilot project was 50.RecruitmentParticipants were recruited through both self- andprovider-referrals between October 16 and December 30,2014. Self-referrals were generated through 1) advertise-ments on the gay social/sexual networking applicationGrindr from October 16 to 23, 2014, 2) advertisements onthe LGBT newspaper/website Xtra from October 16to December 30, 2014, and 3) word of mouth gener-ated by these advertisements. To solicit providerreferrals, two research team members visited 10Toronto-area community-based organizations (CBOs)working in gay men’s health, including a major sexualhealth clinic popular with the city’s gay communityWilton et al. BMC Public Health  (2018) 18:292 Page 2 of 11(Hassle Free Clinic), from mid-October to mid-November, 2014. These visits provided basic PrEP in-formation, an overview of study design, and instruc-tions on how to refer clients to the study. Each CBOwas asked to distribute uniquely numbered referralcards to clients whom they thought would benefitfrom PrEP. The referral cards contained study andreferral information and were used to quantify number ofCBO referrals. Importantly, none of the recruitment mate-rials (advertisements, CBO visits, referral cards) explicitlydescribed the study eligibility criteria related to HIV risk,but simply called for gbMSM at high risk of HIV infection.Interested individuals contacted a single trained researchcoordinator by telephone and the coordinator assessedmen’s eligibility related to age and location only. Thosemeeting age/location criteria were offered an in-personscreening visit.Data collectionMen who attended the screening visit and consented tostudy participation completed a self-administered elec-tronic questionnaire covering sociodemographics, methodof referral, reasons for interest in PrEP, and sexualbehaviours. The questionnaire also included validatedscales to measure HIV risk, syndemic health problemsand attitudes towards sexual identity/behaviours, as de-tailed below.HIV risk was quantified using the 7-item HIRI-MSMscreening tool [9]. This tool was derived from two co-horts of gay men in the United States conducted in thelate 1990s. Scores range from 0 to 47 based on questionsrelated to age, number of male sex partners, number ofcondomless receptive anal sex acts, number of HIV-positive male partners, number of condomless insertiveanal sex acts with HIV-positive partners, use of metham-phetamines, and use of amyl nitrates (“poppers”). Allquestions ask about behaviours in the past 6 months. Ascore ≥ 10 is the suggested cutoff for identifying men athigher risk of HIV infection.We used the Center for Epidemiological Studies - De-pression (CES-D) scale [19] to screen for depressivesymptoms. Scores range from 0 to 60 and, based onprior studies, a score ≥ 23 was considered a positivescreen for a high likelihood of current major depressivedisorder [20, 21]. Cronbach’s α for the CES-D scale inour sample was 0.83.We used the 10-item Alcohol Use Disorders Identi-fication Test (AUDIT) to screen for excessive drink-ing (α = 0.81) [22]. A score ≥ 8 is a recommendedindicator of harmful alcohol use and possible alcoholdependence [23]. We defined multiple substance useas the use ≥2 recreational drugs (methamphetamine,cocaine, crack, ketamine, ecstasy, MDMA, GHB orpoppers) in the last 3 months.We used the Sexual Compulsivity Scale (SCS) toscreen for sexual compulsivity. We defined scores ≥24(α = 0.92) as indicating sexual compulsivity, consistentwith published literature on this construct amongMSM [24–26].We then calculated syndemic count scores [15] byassigning one point each for the presence of depressivesymptoms, harmful alcohol use, multiple substance useor sexual compulsivity, producing values ranging from 0to 4.Statistical analysisThe primary objective was to compare our outcomes ofinterest (HIRI-MSM and syndemic count score) byreferral source. Participants referred to the study by aCBO or an independent physician were classified asprovider-referred, while those whose referral source wasan advertisement, online/social media, or friend/word ofmouth were classified as self-referred.We used Pearson’s chi-square tests (for categoricalvariables) and Wilcoxon rank sum tests (for continuousvariables) to examine whether participant characteristicsdiffered between the two referral groups. In addition, wefit univariable and multivariable regression models withsandwich estimators to explore the association betweenreferral source (independent variable with self-referral asreferent category) and each of our outcomes of interestand their composites (dependent variables). We usedlinear regression for continuous outcomes, Poissonregression for count outcomes and logistic regression forbinomial outcomes. For multivariable models, wecontrolled for race/ethnicity, education and age. Age wasexcluded from models with HIRI-MSM as thedependent variable of interest, as age is a compositecomponent of this scale. For binomial models we usedthe exlogistic command in Stata (StataCorp., 2013) to fitan exact logistic regression model, which produces moreaccurate inferences in small samples than the standardmaximum-likelihood-based logistic regression estimator.Finally, to assess whether the relationship between HIVrisk and referral source varied according to the burdenof syndemic conditions, we conducted a post-hoc ana-lysis in which an interaction term between HIRI-MSMscore and syndemic score was added to a multivariablemodel with referral source as the dependent variable. Allstatistical tests were two-tailed and all analyses wereconducted in Stata 13.1 (StataCorp., 2013).Sample size considerationsTo accrue the target sample size of 50 participants forthe PREPARATORY-5 pilot trial, we anticipated screen-ing 80–100 individuals, with a ratio of self-referrals toprovider referrals between 1 and 1.5. In prior studiesamong Toronto MSM from our group, the standardWilton et al. BMC Public Health  (2018) 18:292 Page 3 of 11deviation of HIRI-MSM scores was 8.6 points [7]. Weestimated [27] that our analysis would be able to de-tect a true difference of roughly 5 points in the meanHIRI-MSM scores between self- and provider-referredparticipants, setting α = 0.05 and power at 80%. Thisis a clinically meaningful difference in HIRI-MSMscores, as it is similar in magnitude to the pointsassigned for many risk factors on the HIRI-MSM tool(eg. 5 points for methamphetamine use, 4 points perincrement in number of HIV-positive sex partners).ResultsDuring the recruitment period, the advertisements onGrindr and Xtra generated 1460 and 58 click-throughs,respectively, while the 10 CBOs referred 115 individualsto the study. Overall, 165 men inquired about the trial,of which 86 underwent screening, were deemed eligible,and consented to study participation (Fig. 1).The median (IQR) age of screened participants was 33(27–40) years. The majority were White/non-Hispanic(72.1%), had a college degree or higher (76.7%), identi-fied as gay (94.2%), and had a primary physician (80.2%)(Table 1). Prior knowledge of PrEP was very high(91.9%) and 14.0% had previously used nPEP. Themajority (60.5%) were self-referrals, primarily throughGrindr, with the remainder (39.5%) referred by pro-viders, primarily a sexual health clinic. A quarter (26.9%)of self-referred men were referred through friends orword of mouth. The most common reasons for studyparticipation were “To protect myself from HIV”(93.0%), “I want to contribute to scientific research”(86.1%), and “To make it safer for me to have sexwithout a condom” (72.1%). There were no significantdifferences in demographic characteristics (Table 1) orreasons for study participation (Fig. 2) by referral source.The median (IQR) HIRI-MSM score of screened menwas high at 26 (19–33), with the vast majority (96.5%)exceeding the recommended threshold (score ≥ 10) fordefining high risk [9]. The majority reported > 10male sex partners (75.3%) and ≥1 condomless receptiveanal sex event(s) (74.2%) in the past six months. Theprevalence of syndemic health problems was also high,with 39.5% meeting criteria for major depressive symp-toms, 39.5% for alcohol related problems, 31.4% for mul-tiple drug use, and 31.4% for sexual compulsivity. Themedian (IQR) syndemic count was 1 (1–2), with 20.9% ofthe sample scoring ≥3.In the primary analyses, there was no difference inHIV risk by referral source (Table 1); the median (IQR)HIRI-MSM score was 26 (19–32.5) for self-referrals and29 (20–34) for provider referrals (p = 0.28). Reporting ofcondomless receptive anal sex in the past 6 months wasalso similar between self- and provider referrals (75.0%vs. 73.5%, p = 0.88). However, popper use in the lastthree months (a component of the HIRI-MSM) wasmore common among provider-referrals compared toself-referrals (76.5% vs. 50.0%, p = 0.01). There was alsoa similar prevalence of all four syndemics in bothFig. 1 Recruitment and referral pathway to the PREPARATORY-5 demonstration project. Participants could be referred through either self orprovider-referred pathways. Specific information on referral source only collected for screened participants. Number of uniquely numberedreferrals cards distributed by CBOs to clients used to quantify CBO referrals. CBO = community-based organizationWilton et al. BMC Public Health  (2018) 18:292 Page 4 of 11Table 1 Characteristics of PREPARATORY-5 participants and comparison by referral sourceTotal(N = 86)Self-referred(N = 52)Provider-referred(N = 34)paReferral Source, n (%) NAAn advertisement on Xtra 4 (4.65) 4 (7.69)An advertisement on Grindr 30 (34.88) 30 (57.69)Facebook/Online/Google 4 (4.65) 4 (7.69)Friend/Word of mouth 14 (16.27) 14 (26.92)Hassle Free Clinic 21 (24.42) 21 (61.76)Other community based organization 8 (9.30) 8 (23.53) Community basedorganizationFamily physician 5 (5.81) 5 (14.71)Age, in years, Median (IQR) 33 (27–40) 32 (27–37) 34.5 (28–42) 0.25Race/ethnicity, n (%) 0.47White, non-Hispanic 62 (72.09) 39 (75.0) 23 (67.65)Non-Whiteb 24 (27.91) 13 (25.0) 11 (32.35)Income, n (%) 0.08Under $20,000 16 (18.60) 7 (13.46) 9 (26.47)$20,000–$39,999 19 (22.09) 11 (21.15) 8 (23.53)$40,000–$59,999 21 (24.42) 16 (30.77) 5 (14.71)$60,000–$79,999 14 (16.28) 10 (19.23) 4 (11.76)$80,000–$99,999 6 (6.98) 1 (1.92) 5 (14.71)Over $100,000 10 (11.63) 7 (13.46) 3 (8.82)Education, n (%) 0.79No college degree 20 (23.26) 13 (25.00) 7 (20.59)College degree or higher 66 (76.74) 39 (75.00) 27 (79.41)Sexual Orientation, n (%) 0.99Gay 81 (94.19) 49 (94.23) 32 (94.12)Bisexual 5 (5.81) 3 (5.77) 2 (5.88)Has a primary care physician, n (%) 69 (80.23) 44 (84.62) 25 (73.53) 0.21Ever heard about PrEP, n (%) 79 (91.86) 48 (92.31) 31 (91.18) 0.99Ever used PrEP, n (%) 12 (13.95) 5 (9.62) 7 (20.59) 0.21Used sexual performance-enhancing drugs, last 3 months, n (%) 31 (36.05) 16 (30.77) 15 (44.12) 0.20Perceived risk (0–100%) of becoming infected with HIV in next year,Median (IQR)37.5 (20–60) 30 (20–52.5) 50 (20–60) 0.36HIRI-MSM and component variablesHIRI-MSM Score, Median (IQR) 26 (19–33) 26 (19–32.5) 28.5 (20–34) 0.28Used poppers, last 3 months, n (%) 52 (60.47) 26 (50.00) 26 (76.47) 0.01Used amphetamine, last 3 months, n (%) 12 (13.95) 8 (15.38) 4 (11.76) 0.76Age< 18 years, n (%) 0 (0.0) 0 (0.0) 0 (0.0) 0.2018–28 years, n (%) 26 (30.23) 17 (32.69) 9 (26.47)29–40 years, n (%) 40 (46.51) 27 (51.92) 13 (38.24)41–48 years, n (%) 13 (15.12) 5 (9.62) 8 (23.53)> 48 years, n (%) 7 (8.14) 3 (5.77) 4 (11.76)Wilton et al. BMC Public Health  (2018) 18:292 Page 5 of 11groups, and an equal median (IQR) syndemic count at 1(0–3) and 1 (1–2) among self and provider-referrals,respectively (p = 0.84).After controlling for other variables in multivariableregression analyses, conclusions related to our uncon-trolled analyses remained unchanged (Table 2). Beingprovider-referred was associated with a greater odds ofpopper use (vs. self-referred, aOR = 3.2, p = 0.03) inmultivariable analysis. No other statistically significantdifferences were detected.Finally, in post-hoc analyses exploring whether thenumber of mental health conditions (syndemic score)modified the relationship between HIV risk and referralsource, we found that the interaction term betweenHIRI-MSM score and syndemic score was not significant(p = 0.75).DiscussionUncertainty exists with regards to the role of CBOs andnon-clinical services and providers in the implementa-tion of “biomedical” interventions such as PrEP [28, 29],but our findings suggest that online strategies and CBOsworking in gay men’s health can play important roles inconnecting individuals at high HIV risk to PrEP services.The rate of gbMSM referrals to our study was high, andsimilar to that observed in open label PrEP studies inthe United States [2, 30]. Contrary to our hypothesis,however, the level of HIV risk and burden of syndemichealth problems were similarly high among gbMSM whoself-referred to this pilot PrEP demonstration project,compared to those who were provider-referred.Men screened for participation in our PrEP demon-stration project were at high risk of HIV infection. Themedian (IQR) HIRI-MSM score of 26 (19–33) amongscreened gbMSM was much higher than in a 2014–2015sample of over 400 gay men testing for HIV at a busysexual health clinic in downtown Toronto (median = 15,IQR = 8–19) [7]. This finding is consistent with otherdemonstration projects showing that PrEP attracts gaymen at highest risk of HIV infection [31, 32]. In the ori-ginal cohorts used to derive the HIRI-MSM, only 8–9%scored 26 or greater, and this threshold was associatedwith a specificity of 92–93% for predicting HIV infectionin the next 6 months [9]. Further, a recent modelingstudy identified a HIRI-MSM score of 25 or more as acost-effective threshold for targeting PrEP to MSM.While this modeling study is subject to the same limita-tions inherent to the HIRI-MSM tool (discussed furtherTable 1 Characteristics of PREPARATORY-5 participants and comparison by referral source (Continued)Total(N = 86)Self-referred(N = 52)Provider-referred(N = 34)paTotal number of male sex partner, last 6 months, n (%) 0.280–5 male partners 7 (8.24) 5 (9.62) 2 (6.06)6–10 male partners 14 (16.47) 11 (21.15) 3 (9.09)> 10 male partners 64 (75.29) 36 (69.23) 28 (84.85)Receptive CAS, last 6 months, n (%) 0.880 22 (25.58) 13 (25.00) 9 (26.47)1 or more times 64 (74.22) 39 (75.00) 25 (73.53)HIV-positive male partner, last 6 months, n (%) 0.25< 1 positive partner 35 (40.70) 23 (44.23) 12 (35.29)1 positive partner 15 (17.44) 11 (21.15) 4 (11.76)> 1 positive partner 36 (41.86) 18 (34.62) 18 (52.94)Insertive CAS, last 6 months, n (%) 0.060–4 times 67 (77.91) 44 (84.62) 23 (67.65)5 or times 19 (22.09) 8 (15.38) 11 (32.35)Syndemic-related factorsSyndemic Count, Median (IQR) 1 (1–2) 1 (0–3) 1 (1–2) 0.84Depressive Symptoms, last week, n (%) 34 (39.53) 22 (42.31) 12 (35.29) 0.51Alcohol related problem, n (%) 34 (39.53) 22 (42.31) 12 (35.29) 0.51Multiple Drug Use, last 3 months, n (%) 27 (31.40) 13 (25.00) 14 (41.18) 0.11Sexual compulsivity, n (%) 27 (31.40) 18 (34.62) 9 (26.47) 0.43PrEP pre-exposure prophylaxis, NANot applicable, CAS condomless anal sex, HIRI-MSM HIV Incidence Risk Index for Men who have sex with men,IQR Interquartile rangeaChi-sq test/Exact test p-values for categorical variables; Wilcoxon rank-sum test p-values for continuous variablesbEast Asian (10.5%), Unidentified (5.8%), Arab/Middle Eastern (4.7%), Black (3.5%), Mixed Race (2.3%), South Asian (1.2%)Wilton et al. BMC Public Health  (2018) 18:292 Page 6 of 11below), it suggests PrEP in Toronto may be cost-effective if the HIV risk profile in future users is as highas in our sample [33].Screened men also had a high burden of syndemics.Approximately one-third were identified as having de-pressive symptoms, an alcohol-related problem, multipledrug use, or sexual compulsivity and one-fifth had threeto four of these problems. This high prevalence is con-sistent with data showing that syndemic health problemsconcentrate in urban gbMSM and synergize to producehigh HIV risk [15, 16, 34], and also highlights the poten-tial to use PrEP programs as a gateway to other healthservices [35]. Another study from our group identified asimilarly high burden of syndemic problems amongpatients accessing nPEP [18], and we have elsewhereargued for the routine implementation of screeningstrategies for such syndemics in all PrEP programs [17].Multiple drug use in our study was similar to the base-line prevalence among gbMSM enrolled in the US PrEPDemonstration Project, where 20.1% reported use of 3or more of poppers, ketamine, ecstasy, gamma-hydroxybutyrate, cocaine, methamphetamine, or erectiledysfunction drugs in the past 3 months [30]. Interven-tions (eg. peer navigators [36]) to link PrEP users torelevant health services (e.g. psychosocial, medical andmental), along with data evaluating the impact of suchservices on PrEP-related outcomes, are urgently needed.To our knowledge, only one other PrEP demonstrationproject has compared screened gbMSM by referralsource. In that study, PrEP was integrated into STDclinics in San Francisco and Miami, as well as a commu-nity health center in Washington, D.C., and self-referredindividuals exhibited greater HIV risk behaviour andgreater perceived HIV risk than those who wereclinic-referred [37]. Further, self-referred men werealso more likely to initiate PrEP [37] and, amongthose who started PrEP, there was a trend towardsbetter adherence (p = 0.07) and retention (p = 0.08)among self-referred men [30]. There may be severalreasons for the discrepancy between this study’s find-ings and our own. First, in the US-based study, par-ticipants initiated PrEP in the same clinic where theclient was identified, while in our study providers re-ferred participants to an external study site. That ourprovider-referred participants had to complete theextra step of attending the study site likely meantthey were more highly motivated to seek PrEP (andthus more similar to self-referrals), potentiallyexplaining the lack of difference by referral type inour study. Second, the “self-referral” category in theUS-based study included individuals who were referred tothe PrEP clinic by their primary care providers; this groupwould have been classified as provider-referred in ourstudy. Third, in our study, self-referrals primarily cameFig. 2 Reasons for wanting to participate in PREPARATORY-5 by referral source. Participants could provide multiple reasons for study participation.Participants were defined as self- or provider-referred based on their primary reported referral pathway, such that some self-referred participantsstill reported “provider suggestion” among their reasons for wanting to join the study. P values calculated using chi-squared analysesWilton et al. BMC Public Health  (2018) 18:292 Page 7 of 11from Grindr and provider-referrals from a popularcommunity-based STI clinic in Toronto, both of whichhave been previously shown to attract high risk gay men[7, 38]. Therefore, it may not be surprising that risk pro-files were similarly high by referral group. However, notall of those frequenting these services are at high risk,and it is reassuring that data from both our studyand the US-based demo project suggest that manygbMSM recognize their elevated HIV risk and poten-tial need for PrEP, and that promoting self-referrals isan important strategy for identifying appropriate PrEPcandidates. Further, our results suggest that Grindrand community-based STI clinics can play importantroles in promoting future PrEP uptake.Many self-referrals in our study were referred byfriends or through word of mouth, suggesting an import-ant role for social networks in improving awareness anduptake of biomedical HIV technologies [39, 40]. Market-ing campaigns designed to promote diffusion of infor-mation through social networks may facilitate morewidespread PrEP self-referrals among individuals at highrisk [41]. Importantly, interventions to promote self-referrals will need to simultaneously address the lack ofknowledgeable providers to which people can referthemselves. To overcome this barrier, our team is cur-rently evaluating a strategy in which patients themselveslink their providers with accredited continuing medicaleducation resources on PrEP [42].Table 2 Regression models exploring the association between referral source (independent variable; provider-referred vs. self-referred)and outcomes of interest (dependent variable)Univariable models Multivariable modelsMOA (95% CI) p MOA (95% CI) pOutcomes of interest (dependent variable)Syndemic scorea IRR = 0.96 (0.66, 1.38) 0.82 aIRR = 1.01 (0.72, 1.41) 0.95Syndemic score compositesaPresence of depressive symptoms, last week OR = 0.75 (0.27, 1.98) 0.67 aOR = 0.74 (0.27, 1.98) 0.67Presence of alcohol related problem, last year OR = 0.75 (0.27, 1.98) 0.67 aOR = 0.91 (0.32, 2.58) 0.99Use of multiple substance, last 3 months OR = 2.08 (0.75, 5.87) 0.18 aOR = 2.20 (0.78, 6.44) 0.15Presence of sexual addiction OR = 0.68(0.23, 1.92) 0.58 aOR = 0.74 (0.24, 2.12) 0.70HIRI-MSM scoreb β = 2.10 (−1.77, 5.96) 0.29 aβ =2.04 (−1.84, 5.92) 0.30HIRI-MSM score compositesbUsed poppers, last 3 months OR = 3.20 (1.14, 9.78) 0.02 aOR = 3.16 (1.08, 10.12) 0.03Used amphetamine, last 3 months OR = 0.74 (0.15, 3.05) 0.89 aOR = 0.69 (0.14, 2.87) 0.82Age (in years) β = 2.74 (−1.53, 7.02) 0.21 aβ = 2.66 (−1.39, 6.72) 0.20Total number of male sex partners, last 6 months≤ 10 male partners Ref. Ref.> 10 male partners OR = 2.46 (0.74, 9.67) 0.17 aOR = 2.35 (0.72, 9.15) 0.19Number of receptive CAS male partners, last 6 months0 times Ref Ref1 or more times OR = 0.93 (0.31, 2.89) 0.99 aOR = 0.93 (0.31, 2.81) 0.99Number of HIV-positive male sex partners, last 6 months< 1 positive partner Ref Ref≥ 1 positive partner OR = 1.45(0.55, 3.94) 0.55 aOR = 1.43 (0.55, 3.87) 0.56Number of insertive CAS HIV-positive male partners, last 6 months0–4 times Ref Ref5 or more times OR = 2.59 (0.82, 8.61) 0.11 aOR = 2.82 (0.88, 9.55) 0.09Measure of association (MOA) in each row is derived from a separate regression model and refers to the association between the referral source (independentvariable; self-referred = referent category) and outcome of interest (independent variable). An OR or IRR of greater than 1 (or a positive β value) indicates thatbeing provider-referred was positively associated with the outcome of interest. For example, being provider-referred was associated with a lower odds ofdepressive symptoms (OR = 0.75), but the finding was not statistically significant (95% CI 0.27–1.98). MOA: Measure of Association; CI: Confidence Interval; β: betacoefficient from linear regression; aβ: adjusted beta coefficient from linear regression; IRR: incident rate ratio from Poisson regression; aIRR: adjusted incident rateratio from Poisson regression; OR: odds ratio from logistic regression; aOR: adjusted odds ratio from logistic regression; HIRI-MSM: HIV Incidence Risk Index forMen who have Sex with Men; CAS: Condomless Anal Sex; CES-D: Center for Epidemiologic Studies Depression Scale; AUDIT: Alcohol Use Disorder IdentificationTest; SCS: Sexual Compulsivity Scale. aMultivariables models are controlled for age, race/ethnicity, and education; bMultivariables models are controlled forrace/ethnicity, and educationWilton et al. BMC Public Health  (2018) 18:292 Page 8 of 11Provider-referrals will also remain important to pro-moting PrEP uptake, particularly for objectively high riskgbMSM who underestimate their HIV risk [7], thoughsuch men may be difficult to reach if they are notengaged in care or “out” to their provider. Although wefound no significant difference in perceived HIV risk byreferral type, our study was underpowered for this com-parison. Expanding continuing medical education onPrEP and the promotion of clear clinical indications forits use could increase provider-initiated PrEP in future.Strengths of our study include being Canada’s firstPrEP demonstration project and our use of validatedscales and concealment of eligibility criteria to minimizereporting bias during participant screening. Our studyalso has limitations that warrant consideration. First, ourmodest sample size may have limited our ability to de-tect small differences between referral sources. Second,because PREPARATORY-5 was the first opportunitywithin Toronto’s gay community to access PrEP at nocost, those seeking study participation may have been“early adopters” and different from the broader gay com-munity. However, the impact of such differences on ourprimary research questions regarding referral sources isunknown. Further, our sample was mostly White, collegeeducated, and previously aware of PrEP – potentiallylimiting generalizability of our results. Differences be-tween provider- and self-referrals may become more ap-parent in more diverse populations, as previous studieshave shown that ethnic minorities and individuals withless education are less likely to self-refer [37] and thatproviders may have racial and other biases in prescribingPrEP [43]. Third, although our findings suggest that pro-viders can accurately identify HIV risk, it was not feas-ible to determine whether formal risk assessments wereactually conducted for provider-referred participants.Finally, we used the HIRI-MSM tool to measure HIVrisk, which has several inherent limitations that ourteam has previously described [44]. In particular, thistool has not been validated in our setting and was de-rived from US-based cohorts of gay men conducted overa decade ago [9]. As such, the index does not includeseveral HIV risk factors that have become important inthe modern context, including viral load and PrEP use.Further, the tool does not include upstream HIV riskfactors, including mental health problems, and may notreflect an individual’s HIV risk at the time of measure-ment or within the short-term future, as the HIRI-MSM is retrospective in nature and HIV risk is dy-namic. Regardless, there were no significant differ-ences by referral source in almost any components ofthe HIRI-MSMrisk score (many of which are commonproxies for sexual HIV risk in other studies of gaymen) or mental health problems which are commonlylinked to higher HIV risk.ConclusionsThe high rate of PrEP referrals and prevalence ofsyndemic problems (which are closely associated withHIV risk) in our study population highlight unmethealth needs in Toronto’s gay community. Our studysuggests that strategies to promote both provider andself-referrals for PrEP are needed to reach at-risk indi-viduals who could benefit most.AbbreviationsAUDIT: Alcohol Use Disorders Identification Test; CBO: Community-basedorganization; CES-D: Center for Epidemiological Studies - Depression;gbMSM: gay, bisexual and other men who have sex with men; HIRI-MSM: HIV Incidence Risk Index for Men who have Sex with Men;IQR: Interquartile range; nPEP: non-occupational post-exposure prophylaxis;PrEP: Pre-exposure prophylaxis; SCS: Sexual Compulsivity ScaleAcknowledgementsWe would like to acknowledge and thank the study participants.FundingThis work was supported by grants from the Ontario HIV Treatment Network(OHTN) and the Canadian Institutes of Health Research (CIHR). DHST issupported by a New Investigator Award from the CIHR/OHTN, and TAH issupported by an OHTN Applied HIV Research Chair.Availability of data and materialsThe datasets used and/or analysed during the current study are availablefrom the corresponding author on reasonable request.Author’s contributionsDHST led and oversaw all aspects of the study; AS and JL conductedparticipant recruitment and data collection; JW and JL visited and presentedstudy information at community-based organizations; DHST, SWN, and JWplanned the study analysis; SWN performed the data analysis; JW, SWN andDHST prepared the first draft of the manuscript. All authors provided inputinto the study design, critically reviewed the manuscript, and approved thefinal version of the manuscript.Ethics approval and consent to participateThe study was approved by the Research Ethics Boards of St. Michael’sHospital, Ryerson University, and the University of Toronto. Informed consentwas obtained prior to the collection of study data.Consent for publicationNot applicable.Competing interestsDHST has received honoraria from Merck, Abbvie, Gilead and Viiv Healthcarefor developing educational lectures; investigator-driven grants from Gileadand Viiv Healthcare; and clinical trial contracts from Gilead and GSK. Theother authors declare that they have no competing interests.Publisher’s NoteSpringer Nature remains neutral with regard to jurisdictional claims inpublished maps and institutional affiliations.Author details1Applied Epidemiology Unit, Ontario HIV Treatment Network, Toronto,Ontario, Canada. 2Department of Psychology, Ryerson University, Toronto,Ontario, Canada. 3Division of Infectious Diseases, St. Michael’s Hospital, 30Bond St, 4CC – Room 4-179, Toronto, ON M5B 1W8, Canada. 4Dalla LanaSchool of Public Health, University of Toronto, Toronto, Ontario, Canada.5British Columbia Centre for Disease Control, Vancouver, BC, Canada.6Department of Medicine, University of British Columbia, Vancouver, BC,Canada. 7Hassle Free Clinic, Toronto, Ontario, Canada. 8ACT (AIDS Committeeof Toronto), Toronto, Ontario, Canada. 9Department of Medicine, Universityof Toronto, Toronto, Ontario, Canada.Wilton et al. BMC Public Health  (2018) 18:292 Page 9 of 11Received: 15 September 2017 Accepted: 19 February 2018References1. Grant RM, Lama JR, Anderson PL, McMahan V, Liu AY, Vargas L, et al.Preexposure chemoprophylaxis for HIV prevention in men who have sexwith men. N Engl J Med. 2010;363(27):2587–99.2. Volk JE, Marcus JL, Phengrasamy T, Blechinger D, Nguyen DP, Follansbee S,et al. No new HIV infections with increasing use of HIV Preexposureprophylaxis in a clinical practice setting. Clin Infect Dis Off Publ Infect DisSoc Am. 2015;61(10):1601–3.3. World Health Organization. Consolidated guidelines on HIV prevention,diagnosis, treatment and care for key populations. 2014. http://www.who.int/hiv/pub/guidelines/keypopulations/en/. Accessed 24 July 2017.4. Centers for Disease Control and Prevention. Preexposure prophylaxis for theprevention of HIV infection in the United States - a clinical practiceguideline. 2014. http://www.cdc.gov/hiv/pdf/PrEPguidelines2014.pdf.Accessed 24 July 2017.5. New York State Department of Health AIDS Institute. Guidance for the useof PrEP. 2015. http://www.hivguidelines.org/prep-for-prevention/guideline/.Accessed 24 July 2017.6. Krakower D, Mayer KH. Engaging healthcare providers to implement HIVpre-exposure prophylaxis. Curr Opin HIV AIDS. 2012;7(6):593–9.7. Wilton J, Kain T, Fowler S, Hart TA, Grennan T, Maxwell J, et al. Use of anHIV-risk screening tool to identify optimal candidates for PrEP scale-upamong men who have sex with men in Toronto, Canada: disconnectbetween objective and subjective HIV risk. J Int AIDS Soc. 2016;19(1):20777.8. Ng BE, Moore D, Michelow W, Hogg R, Gustafson R, Robert W, et al.Relationship between disclosure of same-sex sexual activity to providers,HIV diagnosis and sexual health services for men who have sex with men inVancouver, Canada. Can J public health rev can. Santé Publique.2014;105(3):e186–91.9. Smith DK, Pals SL, Herbst JH, Shinde S, Carey JW. Development of a clinicalscreening index predictive of incident HIV infection among men who havesex with men in the United States. J Acquir Immune Defic Syndr 1999.2012;60(4):421–7.10. Moore DM, Kanters S, Michelow W, Gustafson R, Hogg RS, Kwag M, et al.Implications for HIV prevention programs from a serobehavioural survey ofmen who have sex with men in Vancouver, British Columbia: the ManCountstudy. Can J public health rev can Santé Publique. 2012;103(2):142–6.11. Gallagher T, Link L, Ramos M, Bottger E, Aberg J, Daskalakis D. Self-perception of HIV risk and candidacy for pre-exposure prophylaxis amongmen who have sex with men testing for HIV at commercial sex venues innew York City. LGBT Health. 2014;1(3):218–24.12. Klein H, Tilley DL. Perceptions of HIV risk among internet-using, HIV-negativebarebacking men. Am J Mens Health. 2012 Jul;6(4):280–93.13. Hart TA, James CA, Hagan CMP, Boucher E. HIV optimism and high-risksexual behavior in two cohorts of men who have sex with men.J Assoc Nurses AIDS Care JANAC. 2010 Oct;21(5):439–43.14. Gomez GB, Borquez A, Case KK, Wheelock A, Vassall A, Hankins C. The costand impact of scaling up pre-exposure prophylaxis for HIV prevention: asystematic review of cost-effectiveness modelling studies. PLoS Med.2013;10(3):e1001401.15. Stall R, Mills TC, Williamson J, Hart T, Greenwood G, Paul J, et al. Associationof co-occurring psychosocial health problems and increased vulnerability toHIV/AIDS among urban men who have sex with men. Am J Public Health.2003;93(6):939–42.16. Ferlatte O, Hottes TS, Trussler T, Marchand R. Evidence of a syndemicamong young Canadian gay and bisexual men: uncovering the associationsbetween anti-gay experiences, psychosocial issues, and HIV risk. AIDS Behav.2014;18(7):1256–63.17. Tan DHS, Leon-Carlyle M, Mills R, Moses E, Carvalhal A. Self-administeredscreening for Syndemic mental health problems should be routinelyimplemented among MSM PrEP users. J Gay Lesbian Ment Health.2016;20(1):13–20.18. Tan D, Goddey-Erikefe B, Gough K, Yoong D, Naccarato M, Hart T. Highprevalence of syndemic health problems among patients using HIVpost-exposure prophylaxis: implications for program delivery. Can J InfectDis & Med Microbiol. 2014;25(Suppl A):11.19. Radloff LS. The CES-D scale. A Self-Report Depression Scale for Research inthe General Population Appl Psychol Meas. 1977;1(3):385–401.20. Centers for Disease Control and Prevention. (CDC). Impact of September 11attacks on workers in the vicinity of the world trade center–New York City.MMWR Morb Mortal Wkly Rep. 2002;51:8–10.21. Roberts RE, Lewinsohn PM, Seeley JR. Screening for adolescent depression:a comparison of depression scales. J Am Acad Child Adolesc Psychiatry.1991;30(1):58–66.22. Saunders JB, Aasland OG, Babor TF, de la Fuente JR, Grant M. Developmentof the alcohol use disorders identification test (AUDIT): WHO collaborativeproject on early detection of persons with harmful alcohol consumption–II.Addict Abingdon Engl. 1993;88(6):791–804.23. Allen JP, Litten RZ, Fertig JB, Babor T. A review of research on thealcohol use disorders identification test (AUDIT). Alcohol Clin Exp Res.1997;21(4):613–9.24. Parsons JT, Grov C, Golub SA. Sexual compulsivity, co-occurringpsychosocial health problems, and HIV risk among gay and bisexual men:further evidence of a syndemic. Am J Public Health. 2012;102(1):156–62.25. Kalichman SC, Johnson JR, Adair V, Rompa D, Multhauf K, Kelly JA. Sexualsensation seeking: scale development and predicting AIDS-risk behavioramong homosexually active men. J Pers Assess. 1994;62(3):385–97.26. Kalichman SC, Rompa D. The sexual compulsivity scale: further developmentand use with HIV-positive persons. J Pers Assess. 2001;76(3):379–95.27. Dupont WD, Plummer WD. Power and sample size calculations. A reviewand computer program Control Clin Trials. 1990;11(2):116–28.28. Nguyen V-K, Bajos N, Dubois-Arber F, O’Malley J, Pirkle CM. Remedicalizingan epidemic: from HIV treatment as prevention to HIV treatment isprevention. AIDS. 2011;25(3):291–3.29. Wilton J, Tan D, Maxwell J, Fowler S, Grennan T, Hart T, et al. Exploring theroles of community-based organizations (CBOs) in the rollout of HIV pre-exposure prophylaxis (PrEP) among Toronto gay, bisexual and other MSM.Can J Infect Dis & Med Microbiol. 2015;26(Suppl B):43.30. Liu AY, Cohen SE, Vittinghoff E, Anderson PL, Doblecki-Lewis S, BaconO, et al. Preexposure prophylaxis for HIV infection integrated withmunicipal- and community-based sexual health services. JAMA InternMed. 2016;176(1):75–84.31. Dolling DI, Desai M, McOwan A, Gilson R, Clarke A, Fisher M, et al. Ananalysis of baseline data from the PROUD study: an open-label randomisedtrial of pre-exposure prophylaxis. Trials. 2016;17(1):163.32. Kojima N, Davey DJ, Klausner JD. Pre-exposure prophylaxis for HIV infectionand new sexually transmitted infections among men who have sex withmen. AIDS Lond Engl. 2016;30(14):2251–2.33. Ross EL, Cinti SK, Hutton DW. Implementation and Operational Research: ACost-Effective, Clinically Actionable Strategy for Targeting HIV PreexposureProphylaxis to High-Risk Men Who Have Sex With Men. J Acquir ImmuneDefic Syndr 1999. 2016;72(3):e61–7.34. Tulloch TG, Rotondi NK, Ing S, Myers T, Calzavara LM, Loutfy MR, et al.Retrospective reports of developmental stressors, syndemics, and theirassociation with sexual risk outcomes among gay men. Arch Sex Behav.2015;44(7):1879–89.35. Wilton J, Senn H, Sharma M, Tan DH. Pre-exposure prophylaxis for sexually-acquired HIV risk management: a review. HIVAIDS Auckl NZ. 2015;7:125–36.36. Okeke NL, Ostermann J, Thielman NM. Enhancing linkage and retention inHIV care: a review of interventions for highly resourced and resource-poorsettings. Curr HIV/AIDS Rep. 2014;11(4):376–92.37. Cohen SE, Vittinghoff E, Bacon O, Doblecki-Lewis S, Postle BS, Feaster DJ, etal. High interest in preexposure prophylaxis among men who have sex withmen at risk for HIV infection: baseline data from the US PrEP demonstrationproject. J Acquir Immune Defic Syndr. 2015;68(4):439–48.38. Landovitz RJ, Tseng CH, Weissman M, et al. Epidemiology, sexual riskbehavior, and HIV prevention practices of men who have sex with menusing GRINDR in Los Angeles, California. J Urban Health. 2013;90(4):729–39.39. Silvera R, Stein D, Hutt R, Hagerty R, Daskalakis D, Valentine F, et al. Thedevelopment and implementation of an outreach program to identify acuteand recent HIV infections in new York City. Open AIDS J. 2010;4:76–83.40. Wilton J, Kwag M, Grace D, Jollimore J, Trussler T, Marchand T, et al. Gaymen’s awareness of pooled nucleic acid amplification testing: Examiningdiffusion among networks and influence on testing followingimplementation and promotion in Vancouver, Canada. In: 8th IASConference on HIV Pathogenesis, Treatment & Prevention Abstract Book.Vancouver; 2015. p. 358.41. Wei C, Herrick A, Raymond HF, Anglemyer A, Gerbase A, Noar SM. Socialmarketing interventions to increase HIV/STI testing uptake among men whoWilton et al. BMC Public Health  (2018) 18:292 Page 10 of 11have sex with men and male-to-female transgender women. CochraneDatabase Syst Rev. 2011;9:CD009337.42. Sharma M, Mishra S, Grace D, Knox D, Bayoumi AM, Bogoch I, et al. Animplementation science approach to decentralizing the delivery of HIV pre-eposure prophylaxis through family physicians and sexual heatlh clinicnurses. In: 26th annual Canadian conference on HIV/AIDS research abstracts.Montreal; 2017. p. 127–8.43. Calabrese SK, Earnshaw VA, Underhill K, Hansen NB, Dovidio JF. The impactof patient race on clinical decisions related to prescribing HIV pre-exposureprophylaxis (PrEP): assumptions about sexual risk compensation andimplications for access. AIDS Behav. 2014 Feb;18(2):226–40.44. Wilton J, Mishra S, Tan DHS. Considerations for using the HIRI-MSMscreening tool to identify MSM who would benefit most from PrEP.J Acquir Immune Defic Syndr. 2017 Oct 1;76(2):e58–61.•  We accept pre-submission inquiries •  Our selector tool helps you to find the most relevant journal•  We provide round the clock customer support •  Convenient online submission•  Thorough peer review•  Inclusion in PubMed and all major indexing services •  Maximum visibility for your researchSubmit your manuscript atwww.biomedcentral.com/submitSubmit your next manuscript to BioMed Central and we will help you at every step:Wilton et al. BMC Public Health  (2018) 18:292 Page 11 of 11


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items