Data were drawn from a community-based open longitudinal cohort, An Evaluation of Sex Workers Health Access (AESHA), which initiated recruitment in 2010. This study was developed based on substantial community collaborations with sex work agencies since 2005 , is monitored by a Community Advisory Board of 15+ community agencies, and has included experiential staff (current/former sex workers) since inception. Eligibility includes identifying as a cis or trans woman, being 14 years old or older at enrolment, having exchanged sex for money within the last 30 days, and providing written informed consent. To address the challenges of recruiting stigmatized and hidden populations such as sex workers, time-location sampling was used to recruit participants through daytime and late-night (9 pm–2 am) outreach to outdoor/public sex work locations (e.g., streets, alleys) and indoor sex work venues (e.g., massage parlors, micro-brothels, and out-call locations) across Metro Vancouver, BC. In addition, online recruitment is used to reach sex workers working through online solicitation spaces. Indoor sex work venues and outdoor solicitation spaces (‘strolls’) are identified through community mapping and are updated regularly by the outreach team.
Following informed consent, sex workers completed interview-administered questionnaires at baseline and semi-annually. Interviews were conducted at study offices in Metro Vancouver or a confidential space of their choice (e.g., home, work). The questionnaire is administered by a trained interviewer and data are securely collected and managed using REDCap electronic data capture tools [27, 28]. The questionnaire includes questions related to individual socio-demographic characteristics (e.g., age, sexual and gender identity and orientation, physical and mental health, patterns of substance use), sex work history and patterns, and structural factors . Questions on structural factors included education, racialization, physical and sexual workplace violence, work environment, criminalization, interactions with police, community participation, and access to health and social services (e.g., unmet health needs; barriers to accessing diverse health and social services). The questionnaire is updated regularly in order to capture emerging and changing priorities and needs within the community. All participants received 40 Canadian Dollars (CAD) at each semi-annual visit for their time, expertise and travel.
Sex work community participation variable
Drawing on existing measures of sex work ‘community empowerment ’, our primary exposure variable was a time-updated measure of “sex work community participation” (CP), defined as listing a sex work organization in response to the question “In the last 6 months, have you participated, worked, or volunteered with any community organizations or peer-based initiatives? If yes, which organizations?” Sex work organizations were Metro Vancouver organizations which provide services to sex workers, such as health services and drop-in spaces, many of which are sex worker-led or have implemented peer staff models. Grassroots forms of sex work community participation, such as “spotting” for other sex workers (e.g., tracking client data) and street cleanup, were also captured.
STI Seropositivity outcome
Our primary outcome variable was “STI seropositivity”, which was a time-updated measure defined as a positive STI test (syphilis, chlamydia, or gonorrhea) at each study visit. Following extensive pre-test counselling, participants receive voluntary testing by a project nurse and are offered referrals or STI treatment onsite, if needed, regardless of enrolment in the study. At each semi-annual visit for voluntary sexual health testing, urine samples were collected for gonorrhea and chlamydia, and blood was drawn for syphilis. Syphilis was tested using the rapid plasma reagin (RPR) (97.2% Se; 94.1% Sp) and the Treponema pallidum hemagglutinin assay (TPHA) for all samples with positive RPRs. RPR titers≥1:8 was considered indicative of active infection in the absence of treatment. All participants received post-test counselling. Treatment was provided by a project nurse onsite for symptomatic STIs.
Other explanatory variables
Other explanatory variables of interest were selected a priori based on literature related to “community empowerment” and sex workers’ health and safety. For time-fixed socio-demographic variables, we measured race and Indigenous identity to examine the effects of racism defined as Indigenous (inclusive of First Nations, Métis, or Inuit), Women of Colour (e.g., Black, East Asian, Southeast Asian) vs white. Given the low proportion of participants who identified as Black in our sample, we combined Black and Women of Colour (WOC) for analyses to examine effects of racialized community identities. We also measured high school attainment, im/migration to Canada, sexual minority (lesbian, bisexual, asexual, queer vs. heterosexual), trans women (including transgender women, transexual women and other transfeminine identities- vs cisgender women). Other individual variables were time-updated at each semi-annual study visit, including age (continuous, in years) and years in sex work (continuous). Time-updated variables capturing experiences in the last six months included inconsistent condom use with clients (i.e., using condoms less than 100% of the time for vaginal or anal sex with clients), any non-injection drug use (excluding alcohol and cannabis) and any injection drug use.
Time-updated, structural variables included average weekly income from servicing clients ($CAD), any barriers to receiving health care (e.g., lack of availability, language barriers, poor treatment by health care providers, etc.). We assessed recent (last six months) violence experiences including rape or physical violence by aggressor posing as a client; negative police encounters while working; spent time in jail overnight or longer; and primary place of service, including outdoor/public (e.g. street, public washroom, car, tent), informal indoor (e.g. sauna/steam bath, bar/nightclub, own or client’s place of residence), or formal in-call (e.g. massage/beauty parlor, micro-brothel). This variable included a category for no recent sex work, as not all participants do sex work at every semi-annual visit. Two variables captured place of residence: living in Vancouver’s Downtown Eastside (DTES) and living within the City of Vancouver. The DTES represents an inner-city neighborhood within the City of Vancouver’s downtown core where community organizations and low-barrier services are heavily concentrated.
Baseline descriptive statistics for individual and structural characteristics were calculated as frequencies and proportions for categorical variables and measures of central tendencies (i.e., mean, median, interquartile range (IQR)) for continuous variables. These were stratified by the primary dependent variable of interest and compared using Pearson’s chi-square test for categorical variables (in the case of small cell counts, Fisher’s exact test was used in place of Pearson’s chi-square) and the Wilcoxon rank-sum test for continuous variables.
Bivariate and multivariable analyses used logistic regression with generalized estimating equations (GEE) and an exchangeable correlation matrix to account for repeated measurements amongst participants over time . Bivariate analyses examined associations with community participation over the study period. Explanatory variables significantly associated with community participation at p < 0.10 were considered for inclusion in the multivariable explanatory model. The multivariable model with the best overall fit, indicated by the lowest quasi-likelihood under the independence model criterion, was determined using a backward selection process. Finally, we developed a separate multivariable GEE confounder model to identify the independent association of community participation on STI seropositivity. All variables from the full explanatory model were considered as potential confounders and included sexual minority, trans identity, racialization, im/migration status, living in the City of Vancouver, rape, and weekly income from servicing clients. Statistical analyses were performed in SAS version 9.4 (SAS, Cary, NC), and all p-values are two-sided.