Data for this study derives from the baseline, an online survey of a randomized controlled trial (RCT). The aim of the RCT was to investigate the effect of a comprehensive crowdsourced intervention to increase HIV testing uptake among Chinese MSM. A detailed description of the study design was provided in the published protocol (26). This baseline survey was conducted between June and August 2016 in eight cities in China (Guangzhou, Shenzhen, Zhuhai, Jiangmen, Jinan, Qingdao, Jining, and Yantai). Participants were recruited through banner advertisements on Blued (Blue Brother, Beijing, China), the largest mobile phone application for MSM social networking in China, with over 40 million users worldwide (27). MSM, who clicked the survey link was directed to the information page and survey, hosted by SoJump (Sojump, Shanghai, China). Before commencing the survey, we obtained online consent from each participant. Eligibility criteria for the study included being born biologically male, age 16-years or older, currently living in one of the eight study cities, never being diagnosed with HIV infection, and ever had oral or anal sex with a man. Participants’ mobile phone numbers were verified by the survey platform to avoid multiple answers by the same participant. After finishing the survey, each participant received 50 Chinese Yuan (approximately 8 USD) as a mobile phone credit.
Anticipated HIV stigma in this study was assessed through a seven-item questionnaire, developed by Golub and Gamarel (15) for a study among MSM in New York. Survey items included the expectations of discrimination, prejudice from others, and their feelings if they were infected. (Table 2) All the items were rated on a four-point Likert scale (1=Strongly Disagree; 2=Disagree; 3=Agree; 4=Strongly Agree). The mean of the seven-item scale was calculated to create a final anticipated HIV stigma score, which ranged between 1 to 4. A higher score indicated a higher anticipated HIV stigma. Scale reliability was high in this survey (Cronbach’s α = 0.85). The Chinese translation of the questionnaire was piloted with 15 MSM from two study cities, and amendments were made based on their feedback.
Sociodemographic information, including age, education level, income, marital status, gender identity, and self-identified sexual orientation, were collected from each eligible participant. We dichotomized the educational level at high school and annual income at 9200 USD as they were the mean educational and income level of urban China (28). Self-identified sexual orientation was classified as gay or other.
We also collect behavioral information including usage of social media sexual partners seeking in the past 12 months, ever had sex in the previous three months, any condomless sex with male sexual partner(s), and having more than one male sexual partner in the previous three months. We also assessed social media usage (ever vs n.ever) of both gay apps such as Blued and Grindr, and non-gay apps such as Wechat, QQ (both mobile text and voice messaging communication services in China) and Weibo (a China-based social networking and microblogging service websites, similar to Twitter). Testing and health care utilization behaviors were also reported by participants, including ever tested for HIV, ever self-tested for HIV, ever facility tested for HIV, ever tested for HIV in the past three months, and ever utilized any public sexual health service in the previous 12 months. Public sexual health services included: free condoms and lubricant, peer-led sexual education, HIV and STI screening and treatment, pamphlets on HIV/STI-related information, and medical treatment in the public medical facility.
We collected data about levels of gay-community engagement in sexual health and MSM status disclosure of the participants. We used a six-item scale to measure participants, the level of gay-community engagement, and categorized community engagement into four categories, ranging from no engagement to substantial engagement (see the details questions in the appendix) (29). MSM status disclosure was defined as having discussed sexual orientation or MSM sexual history with a health professional or others.
Descriptive analysis of socio-demographic, sexual behavior, and HIV/STI testing related variables were conducted by reporting distribution frequencies among the survey participants. Bivariable and multivariable generalized linear models were used to assess measures of association between anticipated HIV stigma with sexual behaviors, testing, and health care utilization behaviors, and community engagement variables. Sociodemographic characteristics, including age, education, income, sexual orientation, and marital status, and city, were adjusted in the multivariable generalized linear models. All analyses were conducted in SPSS version 22.0 (Armonk, NY, UCA). In the model, we defined statistical significance as p<0.05.
Eligible survey participants were invited to take part in the survey after indicating their informed consent online. The study protocol was approved by the ethics review committees at the Guangdong Provincial Centre for Skin Diseases and STI Control, the University of North Carolina at Chapel Hill, and the University of California, San Francisco.