Two internet-based approaches to promoting HIV counselling and testing for MSM in China

Matt Avery, Gang Meng & Stephen Mills

Published Online: July 17, 2014
Full Text: HTML, PDF


The internet is an increasingly popular among gay men and other men who have sex with men (MSM) in China for finding sexual partners. Gay men and other MSM who meet online are at high risk for HIV infection, but less likely to visit ‘traditional’ venues where they can receive interpersonal HIV prevention interventions. New virtual models are needed to provide HIV prevention messages and services to these gay men and other MSM. FHI 360 and Guangzhou Tongzhi (GZTZ) piloted separate, but complementary, approaches to using information and communications technology to promote uptake of HIV counselling and testing (HCT) among gay men and other MSM in three Chinese provinces (Yunnan, Guangxi and Guangzhou). These approaches included dedicated websites featuring online risk assessment and appointment making, crowd-sourced service promotion messages and dissemination via participants’ microblog accounts and social media profiles. Reach was measured using Web analytics and traditional monitoring and evaluation tools, and government partners provided data on HCT uptake. The FHI 360 and GZTZ interventions reached 7,000 and 2.3 million unique visitors, respectively, and contributed to increases in HCT uptake of 26% and 66% as well as to higher rates of HIV case finding. Internet-based interventions like those conducted by FHI 360 and GZTZ represent a promising channel for engaging otherwise difficult-to-reach gay men and other MSM in China.

Keywords: HIV, men that have sex with men, MSM, ICT, HIV Counselling and Testing, China


A 2008-2009 survey of 61 Chinese cities indicated a nationwide HIV prevalence of 5% among men who have sex with men (MSM),  though specific provinces reported notably higher prevalence, including Yunnan province at 10.9% (Wu, et al., 2013). Since that study was conducted, the China Ministry of Health in 2011 estimated prevalence among this population to be 6.3%, suggesting the epidemic continues to expand (China Ministry of Health, 2012). HIV testing and counselling (HTC) is a key entry point into the cascade of prevention, treatment, care and support for people living with HIV(Hull, Wu, & Montaner, 2012; Kilmarx & Mutasa-Apollo, 2013; Sullivan, et al., 2012); however, increasing HIV prevalence in the MSM population has not resulted in sufficiently increased rates of HIV testing among Chinese MSM. Despite HIV counselling and testing (HCT) services being provided free of charge by the Chinese government, as of 2011 half of Chinese MSM had not received an HIV test within the previous 12 months(China Ministry of Health, 2012). Specific barriers to increasing HTC uptake among Chinese MSM include perceptions that available services are low-quality and discriminatory(USAID/Health Policy Initiative, 2009; Yu, An, & Tong, 2009).Testing behaviour among MSM is also influenced by community norms –a 2010 behavioural  survey among MSM indicated that individuals were more likely to get tested themselves if they perceived testing as a norm among their peers (Population Services International (China), 2010).

Information and communications technology (ICT) platforms including websites, social media and microblogs, are one channel for promoting HTC services since the Internet has become an increasingly popular means of finding sexual partners for MSM in China as well as globally(Lim, Guadamuz, Wei, Chan, & Koe, 2012; Zhang, et al., 2007). Reaching these men with HIV prevention services may be particularly important as studies conducted in China and elsewhere have suggested that MSM who seek sexual partners online may be at higher risk for HIV infection due to a greater likelihood of engaging in unprotected anal sex(Berg, Tikkanen, & Ross, 2013; Grov, 2012; Parsons, Vial, Starks, & Golub, 2013; White, Mimiaga, Reisner, & Mayer, 2013; Zhang, et al., 2007), higher rates of sexually transmitted infections(Lau, Kim, Lau, & Tsui, 2003), or a greater likelihood of having multiple and concurrent sexual partners(Chew Ng, et al., 2013; Li, et al., 2012; Rosser, Miner, et al., 2009; Rosser, Oakes, et al., 2009; Young, Szekeres, & Coates, 2013; Zhang, et al., 2007). While the 61-city survey in China did not find a specific link between Internet use and HIV prevalence, that study did indicate that MSM who interact primarily online are likely to be younger and better-educated than other MSM – the authors suggested that the young age of Internet users could mask undetected, acute infections (Wu, et al., 2013). Further, many MSM who use the Internet to “cruise” for partners never visit or seek sexual partners in traditional gay venues(Saxton, Dickson, & Hughes, 2013). The Internet may thus provide a medium to gain access to a subpopulation of MSM who are at especially high risk, do not necessarily have strong social networks with the local gay community, and are thus not reached by traditional, venue-based peer outreach activities.


Several organisations working in China have piloted what is referred to as an “online-to-offline (O2O)” model where populations of MSM are targeted over web-based platforms where they interact, with the goal of initially engaging these men online in order to foster eventual in-person interaction, including uptake of HIV counselling and testing and sexual health services. In this paper, we present and compare two specific approaches to social media strategies, their evaluation designs and metrics on reach and effectiveness, and options for the future.

Social and Antisocial Media: Two ICT approaches

Yunnan and Guangxi provinces (pop. 46.31 million and 46.45 million, respectively), in southwest China, are among the highest HIV prevalence provinces in the country, with a number of community-based organisations conducting HIV prevention education, HTC referral and community-based testing in both provinces. The six-month “Xiu Boy” campaign was launched by the USAID-funded Spring Rain and Green City Rainbow MSM community-based organisationss in the provincial capitals of Kunming and Nanning, respectively, in order to increase MSM dialogue around and uptake of HTC services. The centrepiece of Xiu Boy was a microsite which hosted HTC information; an online, anonymous HIV risk calculator; and a “digital video” competition wherein participants shared videos of themselves talking about testing across their social media networks (SinaWeibo,, and among others) and encouraged their friends to vote for their favourite video. The campaign was additionally integrated into traditional outreach programming, with trained peer educators promoting the Xiu Boy microsite at MSM “hot spots” including bars, bathhouses and public parks and organizing special campaign events including a launch party and “Show Your Best Self” underwear show.

Guangzhou (pop. 12.78 million), the capital of Guangdong Province in southern China, is roughly twice the size of Kunming or Nanning, with a larger and more cosmopolitan MSM community. Guangzhou Tongzhi (GZTZ) has operated an LGBT-themed website since 1998, and has partnered with the Guangzhou CDC to offer community-based HTC since 2008. In contrast to the Xiu Boy campaign, which integrated in-person interaction and encouraged open experience sharing via social media, GZTZ built an ICT platform which consciously limits the human interaction necessary to promote HIV testing: online games encourage self-efficacy and responsibility, a self-risk evaluation targets awareness of personal-risk and decision-making, online ads publicise services and user-friendly tools facilitate appointment making and deliver testing reminders.
While these interventions took different approaches to harnessing web-based platforms for HCT promotion, they also shared several points in common. Both combined innovative web-based approaches with more traditional service promotion activities; both tied service promotion to service delivery through specific partner agencies, and both partnered with local government (municipal centre for disease control and prevention) to deliver these services.

Evaluation designs and metrics

The two social media projects utilised a variety of designs and metrics to measure their penetration into MSM networks and to estimate their impact on service utilisation. Both used Google Analytics to track data on website usage (site visits, site visitors, % new visits, page views and bounce rate).

Users were directed to the Xiu Boy online risk calculator either via the website home page or else via paid banner ads, displayed on a number of for-profit gay Chinese websites, which linked directly to the calculator. The calculator collected data on website users’ self-identified gender, partners’ gender, number of sexual partners within the last six months, and specific risk behaviours (sex with strangers; commercial sex; sex while under the influence of drugs or alcohol; injection drug use; unprotected oral, anal and vaginal sex) and health-seeking behaviours (STI screening and HIV testing). Frequencies were calculated for these measures using SPSS (Version 11.0), in order to build a risk profile of website users who completed the risk calculator.

Levels of risk as reported by the risk calculator were based on a simple calculation of the number of risk or health-seeking behaviours in which an individual reported engaging. Each individual behaviour (male-to-male sex, multiple sexual partners, sex with unknown partners, commercial sex, STI check-up, sex under the influence of drugs or alcohol, needle sharing, HIV testing, unprotected anal or vaginal sex) was assigned a point value (0-15 points) with particularly risky behaviours (commercial sex, needle-sharing, unprotected vaginal/anal sex) automatically assigned higher values. Final summative scores of 0-2 points were considered minimal risk, 3-4 were considered somewhat risky, 5-14 were considered of moderate risk, and scores of 15 or above were considered high risk. This methodology was adapted from similar risk calculators used in other HIV interventions; however, it was primarily intended to generate increased risk perception among campaign participants and not intended to accurately reflect statistical levels of HIV risk.

Use of the risk calculator was tracked via I.P. address. In order to avoid double-counting of respondents, the database was screened for multiple instance of the same I.P. address. In all cases of multiple entries from the same I.P. address, the entry with the earliest time stamp was retained and all others were removed from the data set. This was based on the assumption that a website user was most likely to respond to risk calculator questions accurately on their first completion of the survey, and then experiment on subsequent attempts with changing their responses to see how it affected the results of their risk calculation.

The digital video competition was evaluated according to the number of videos posted by website users and the number of “Likes” recorded by the website for each video, as tracked via unique I.P. address. Only the first instance of a unique I.P. address was recorded for the purpose of calculating total number of Likes.

The number of individual MSM reached by the Xiu Boy campaign with campaign messages via face-to-face interaction with a project-trained peer educator was tracked using standardised monitoring and evaluation data collection forms. In order to avoid double counting of project clients reached by multiple peer educators, or by the same peer educator multiple times, these data distinguish between “new” and “follow-up” contacts using the recall method so that the total number of persons “reached” is specific to the individual service rendered and does not mix new and repeat clients. GZTZ did not conduct traditional, face-to-face “outreach” activities.

For both interventions, the number of MSM who accessed HIV testing services at an affiliated testing site and received their test result, and the number of MSM tested positive through campaign-affiliated testing sites was recorded through standardised data collection forms. GZTZ was also able to collect data on the number of individuals confirmed positive through Western Blot confirmatory testing, and the number of positive individuals successfully referred to follow-up care, through their partnership with the Guangzhou Municipal CDC. However, antiretroviral treatment (ART) in China is managed through a separate (non-CDC) hospital system; thus, data on ART initiation and maintenance are not reported for clients referred for treatment through these interventions.


Using the above-described metrics, we present the most recently available data and indicators of both the Xui Boy and GZTZ websites and approaches in Table 1 (below).

Table 1. Usage statistics for Xiu Boy and GZTZ websites.

Intervention Site Visits Unique Site Visitors Page Views Bounce Rate
Xiu Boy
(April-Sept. 2011)
9,461 7,082 40,566 53.45%
GZTZ (Jan.-Dec. 2012) 6,679,707 2,298,808 48,899,134 33.67%

For the purposes of these interventions, “site visits” refers to the total number of visits to the specified website, and “page views” refers to the total number of times individual pages within the website were visited. “Unique site visitors” refers to the number of unique individuals who spent time on any page of the website, whether they did so once or multiple times, though this information is subject to data collection errors. The number of unique site visitors is tracked by Google Analytics using “cookies” – small pieces of information installed on a computer when it visits a website, allowing the website to recognise that computer on subsequent visits. If a returning website visitor deletes the cookies stored on their computer, or uses a different machine or Internet browser to visit the site, they may be misclassified as a new unique visitor; thus, Google Analytics tends to place more importance on total site visits.

“Bounce rate”, finally, represents the number of visits when users leave your site after just one page, regardless of how they got to your site or how long they stayed on that page. There are a number of potential explanations for a high bounce rate, including that visitors received the information they needed after visiting only one page, that they visited the site in error or were not interested in the website content, or that they experienced design or usability issues with the website.

XIU BOY campaign

The Xiu Boy campaign ran for six months (April through September 2011) and the social media digital video competition was conducted for the 2nd half of the campaign. During the campaign period there were a total of 9, 461 site visits, with 7,082 unique site visitors and 40,566 page views. The most popular pages by page views, outside of the main landing page, were Information for HIV-positive Individuals, Online Risk Calculator, and Information on Finding a Testing Centre. These figures do not include visitors to the separate web page that hosted the digital video competition.

74.47% of visits to the Xiu Boy website were new visits, and the average user visited roughly 4 pages per visit. The bounce rate (percentage of visitors who enter the site and “bounce” – leave the site – rather than continue viewing other pages within the same site) was 53.45%.

99% of site visitors were from China. Of those visitors, 57% were from the targeted campaign cities, and 70% came from the target provinces. Traffic from outside of the target cities is also significant as MSM from the countryside commonly travel to the provincial capital to access healthcare and other services.

During the campaign period, trained peer educators additionally reached 1,799 MSM through either one-on-one outreach, small group activities, or large-scale community events. While outreach activities were not all specifically related to the Xiu Boy campaign, peer educators were trained to provide campaign messages and promote the website through all outreach encounters. It was not possible to estimate what percentage of individuals reached with one-on-one outreach were also reached via the campaign website.

Online Risk Calculator.961 site visitors accessed the online risk calculator and 904 (94%) completed all items. Of those who completed the calculator, based on their answers to the survey items, 88.9% were at medium or high risk for HIV infection.

Table Two. Xiu Boy Campaign: Characteristics of website visitors who used the anonymous risk calculator (n=961)
n (%)
Gender (n=961)
Male 948 (98.6%)
Female 6 (0.6%)
Transgender 7 (0.7%)
Partners’ gender (n=912)
Male 703 (77.1%)
Female 48 (5.3%)
Both 161 (17.7%)
No. Sex Partners Last 6 Months (n=876)
0 108 (12.3%)
1 235 (26.8%)
2-4 376 (42.9%)
5-10 102 (11.6%)
>10 55 (6.3%)
Sex with a Partner You Do Not Know (n=862)
No 221 (25.6%)
Yes 641 (74.4%)
Engaged in Commercial Sex (n=848)
No 666 (78.5%)
Yes 182 (21.5%)
Sex under the Influence of Drugs or Alcohol (n=831)
No 712 (85.7%)
Yes 119 (14.3%)
Shared Injecting Equipment to Use Drugs (n=824)
No 818 (99.3%)
Yes 6 (0.7%)
Tested for STIs (n=841)
Yes (w/in last 6 months) 145 (17.2%)
Yes (not w/in last 6 months) 185 (22%)
Never Tested 511 (60.8%)
Tested for HIV (n=812)
Yes (w/in last 6 months) 133 (16.4%)
Yes (not w/in last 6 months) 179 (22%)
Never Tested 500 (61.6%)
Sexual behaviour
Oral Sex
Performed Oral Sex on Partners 650 (67.6%)
Consistently Used Condoms when Performing Oral Sex 32 (4.9%)
Received Oral Sex from Partners 657 (68.4%)
Consistently Used Condoms when Receiving Oral Sex 26 (4%)
Anal Sex
Penetrated Partner Anally 499 (51.9%)
Consistently Used Condoms when Penetrating Partner Anally 177 (35.5%)
Penetrated Anally by Partner 474 (49.3%)
Consistently Used Condoms when Being Penetrated Anally 168 (35.4%)
Vaginal Sex
Penetrated Partner Vaginally 115 (12%)
Consistently Used Condoms when Penetrating Partner Vaginally 30 (26.1%)
Risk Profile (n=904)
Minimal Risk 54 (6%)
Some Risk 47 (5.2%)
Medium Risk 317 (35.1%)
High Risk 486 (53.8%)

The majority of risk calculator users self-reported as males (98.6%, n=948) who only had sex with other men (77.1%, n=703). Among those website users who reported engaging in anal sex, 35.5% (n=177) reported consistent condom use as the penetrating partner and 35.4% (n=168) reported consistent condom use as the penetrated partner. A significant minority of users (17.7%, n=161) reported sex with male and female partners, and among those who reported penetrating their partner vaginally only 26.1% reported using condoms consistently.

The majority of users (54.5%, n=478) reported between 2-10 sexual partners within the last six months, and 74.4% (n=641) reported having sex with partners they did not know.

Despite high levels of reported sexual activity and relatively low levels of consistent condom use, 60.8% (n=511) of risk calculator users reported having never been screened for sexually transmitted infections, and 500 (61.6%) had never been tested for HIV (22% had been tested, but not within the last year).

Digital Video Competition

In total, 48 videos were uploaded for the digital video competition and 6,673 total votes were cast – voting was only permitted during the final month of the campaign to avoid privileging videos which were posted earlier. The winning video collected 1,745 votes while the first runner-up received 1,347 votes.

Due to a technical error, page views attributed to the digital video competition were not included in the total views or visitor counts for the Xiu Boy website, which might otherwise have contributed significantly to increasing those numbers as social media viewers could link directly from a video to the competition page. However, the video competition still drove increased traffic to other content on the main Xiu Boy site – during the competition the site recorded just over 6,000 visits (4,466 unique visitors) an increase of 128% over the three-month period preceding the competition.

HIV Counseling and Testing

During the Xiu Boy campaign period, HCT uptake by MSM for the three affiliated clinical sites increased by 26% (from 896 to 1135) and the number of positive cases identified increased by 22% (from 57 to 70) when compared to the previous six-month period. As can be seen in Chart 1 (below), most of the increase in HCT uptake is attributable to the Guangxi campaign site (Site 3), which accounted for 82% of all testing conducted during the campaign period, a 33% increase over the previous six-month period as compared to the 3% increase in testing uptake in Yunnan (sites 1 and 2, combined).

Table 3.HIV testing uptake during the Xiu Boy campaign.

FY11 Q1-Q2 FY11 Q3-Q4
Tested/Tested + Tested/Tested+
Site One 64/13 58/8
Site Two 136/6 148/9
Site Three 696/38 929/53
Total 896/57 1,135/70


An O2O procedure has been developed by GZTZ since 2008 to support MSM in accepting HIV tests and relevant services. Its main components include:

  • An online service module aimed at encouraging, reinforcing and validating safer sex behaviours and awareness of HIV testing by providing information on HIV prevalence and service promotion, offering a self-service risk evaluation tool, and conducting embedded vignette-based interventions;
  • A link between online and offline services (testing and results delivery), to mitigate clients’ unwillingness to receive an HIV test and to advocate for testing among sex partners of newly diagnosed HIV-positive clients by maintaining an online appointment and notification system, self-service results query, and anonymous partner notification;
  • An offline services module to boost clients’ confidence regarding service quality through provision of CBO-based pre- and post-test counseling, rapid testing, and supportive services for HIV-positive individuals.

During calendar year 2012, the GZTZ website received a total of 6,679,707 visits, made by 298,808 unique visitors, who made 48,899,134 page views. Not only did the site (which was more established than the Xiu Boy site) generate significantly more traffic, the bounce rate of 33.67% was noticeably lower, indicating that a higher percentage of those visiting the site actually intended to do so.

Over the course of 2012, GZTZ conducted 5,389 HIV screening tests for MSM, an increase of 130.5% since 2010. Of those MSM tested, 8.57% (n=462) were screened positive, which reflects a 33.76% increase in the 2010 case finding rate (5.1%, n=119). The Guangzhou CDC testing algorithm for members of high-risk populations is a single, rapid HIV screening test followed by Western Blot confirmatory testing which is processed off-site by the CDC. Of those individuals screened positive through the GZTZ service, 90% (n=416) agreed to receive an HIV confirmatory test, and 75% (n=312) were notified of a confirmed positive result. Of the 25% of clients screened positive who did not receive a confirmatory test result:

  • 5.5% (n=23) were found through ID tracking to have already been confirmed positive and thus not re-tested,
  • 2.4% (n=10) received a confirmatory test but were later determined to have been previously confirmed HIV-positive and, thus, not recounted as “new cases”
  • .48% (n=2) were found to be false positives,
  • 16.6% (n=69) did not receive their confirmatory test results within the reporting period, though they may have been informed of their test results at a later date.

CD4 testing is also provided by the CDC (and ART treatment centre) for all confirmed HIV-positive individuals, regardless of where they received their HIV test. Among those GZTZ clients who did receive their confirmatory test result, 12.5% were reported by the Chinese government not to have received a follow-up CD4 test.

Table 3. HIV testing and referral to care and treatment by GZTZ

2012 2011 2010
Received HIV screening (3 affiliated sites) 5,389 3,247 2,338
Screened positive 462 275 119
Agreed to receive conformatory tests 416 N/A N/A
Non-duplicate clients who received confirmatory test 383 179 73
Confirmed to be newly found HIV+ 312 166 59
HIV+ & received CD4 count tests 273 N/A N/A

Screen Shot 2014-07-17 at 00.27.54

Additionally, while there are a number of different community-based organisations that partner with the Guangzhou CDC to offer HIV counselling and testing for MSM (see Figure 1, below), in calendar year 2012 GZTZ was responsible for roughly 83% of all clients screened for HIV within Guangzhou City who identified themselves as MSM.

Screen Shot 2014-07-17 at 00.27.27

Figure 1. MSM screened for HIV in Guangzhou City, 2012


Findings and results from both social media approaches support the hypothesis that web-based platforms can be an effective channel for the promotion of HTC services for MSM in China. Internet-based approaches in China have typically recreated traditional, venue-based outreach practices – the development and distribution of promotional materials, peer education etc. – and recreated these approaches while treating websites and chat rooms as a kind of “virtual venue.” While this approach has its advantages in that it is possible to apply existing tools and manpower with limited need for adaptation or training, the Internet can be a more efficient tool when interventions make use of the unique advantages of this medium. Both approaches described in this paper took different approaches which limit the need for a trained cadre of semi-professional peer staff. GZTZ sought to automate intervention services in order to minimise or completely remove the need for direct human involvement – for instance, a telephone hotline was originally established to help online staff interact with clients; however, calls to the hotline decreased dramatically once dedicated software was put into place allowing users to make appointments and receive transportation directions and service reminders.

In contrast, the Xiu Boy campaign sought to supplement the provision of information about HIV transmission and prevention delivered by trained and employed peer educators with positive normative statements about the desirability of knowing one’s HIV status, which were disseminated by regular service clients and propagated virally through their online social networks. The Xiu Boy microsite additionally allowed potential (but hesitant) clients to view photographs and digitally-recorded introductions of clinic staff, facilities and procedures as well as client testimonials about participating HCT service centres without the need to actually visit a clinic, thus reducing what are sometimes termed the “entry costs” for counselling and testing – fear, discomfort, embarrassment – without the need for direct human intervention. Critically, information collected through the Xiu Boy risk assessment tool indicates that testing information delivered via the website reached a population of men who have never (or not recently) received an HIV test despite engaging in unprotected anal sex either as the penetrating or penetrative partner.

These interventions also demonstrate that evaluation is essential to the success of any public health intervention, and freely available tools like Google Analytics can greatly simplify the process of collecting and analysing monitoring data for web-based intervention activities. However, the greatest obstacle to effective monitoring and evaluation isn’t collecting extensive user engagement data such as the number of “Likes” or the website bounce rates; the key obstacle is a failure to determine what return on investment the online strategy is intended to generate in the first place. It may be tempting to envision web-based metrics as an end unto themselves, but this approach fails to acknowledge the reach and impact of broader communications activities (Gordon, 2013). Rather than thinking in terms of internet metrics versus traditional public health indicators, monitoring and evaluation systems should focus holistically on the overall goal (in this case, HCT uptake) and then identify the best indicators to determine whether intervention strategies are working.

A key barrier to integrating Internet metrics more holistically into monitoring and evaluation  frameworks, however, is the difficulty in tracking unique contacts across virtual and physical engagements, and the primacy in many monitoring and evaluation frameworks (for instance, the UNGASS and PEPFAR indicators) for real-life contact. More research is needed to determine how “virtual” contacts and measures of engagement can best be integrated into these frameworks so as to highlight the increasing relevance and importance of digital interactions.

Coordination between community-based organisations and the municipal government was critical to the success of the GZTZ and Xiu Boy interventions. Across the Asia-Pacific region it is estimated that as many as half of all members of key affected populations are unaware of their HIV status (UNAIDS, 2012),and while community-based HTC is recommended by the WHO (WHO, 2012) and has been demonstrated to be highly acceptable to testing clients (Suthar, et al., 2013), there are numerous policy barriers to adopting this strategy across the region, including in China. The interventions reviewed here contributed to increased service uptake without significant loss to follow-up, false positivity or reported adverse events (breaches of confidentiality, etc.) and thus further demonstrate the potential key role CBOs can play as partners to the China CDC in increasing the number of MSM who know their status and access care and treatment in line with the national strategy.

Two key links in the cascade of HIV prevention to care and treatment which did exhibit worrying loss to follow-up were in the gaps between clients screening positive for HIV infection, receiving a confirmed positive result, and receiving a CD4 test for assessing ART readiness. The current Chinese testing algorithm calls for expensive and technically demanding Western Blot confirmatory testing which typically delays provision of results by one to two weeks, but in some cases more than a month, delaying access to pre-ART staging and treatment initiation and potentially contributing to loss to follow-up. Delays in administering CD4 tests may be an additional barrier to treatment initiation. It is necessary to advocate for the adoption of newer confirmation strategies which would reduce or eliminate wait times for test results (Styer, Sullivan, & Parker, 2011)such as scaling up the use of point-of-care CD4 tests in selected community-based testing centre settings (Jani, et al., 2011; Mtapuri-Zinyowera, et al., 2010), which would help to move clients more efficiently through the HIV care continuum.

Finally, the interventions reviewed here also demonstrate that many people underestimate the human resources and skills needed to develop and sustain a technology- driven intervention. It is worth noting that, of the two intervention sites participating in the Xiu Boy campaign, the Guangxi site far outperformed the Yunnan site terms of testing uptake. This disparity is unlikely the result of differences in service delivery model (both sites offered community-based rapid testing, while only Yunnan additionally offered clinic-based testing) or HIV risk (reported HIV prevalence is in fact significantly higher in Yunnan than Guangxi). However, the implementing team in Guangxi was younger and more social media savvy, as evidenced by team members having their own, pre-existing social media accounts and having significantly more success attracting social media followers and soliciting submissions for the digital video competition. Guangxi additionally accounted for 45.12% of all visits to the Xiu Boy website during the campaign period, as compared to 23.8% from Yunnan.

As the experience of these two programs shows, launching and supporting an Internet-based intervention requires a Web- and social media-savvy communications team who are comfortable working within the platforms the intervention will target, project managers who understand the possibilities and limitations of these technologies, community experts who are in tune with community needs and preferences, administrators who can maintain strong control over the workflow, and a well-trained team of service providers. The unique skill sets needed to design, manage and monitor Web-based and social media activities may not always be available within one organisations but will require several who bring distinct expertise to a well-planned consortium.


There are a number of limitations which should be taken into account when interpreting the above results: most notably that, as neither intervention used a rigorous evaluation design, the influence of confounding variables cannot be discounted when considering the demonstrated increase in HIV testing rates. Potential confounders may include other HIV prevention and test promotion activities (peer education, media coverage etc.) taking place concurrent with the GZTZ and Xiu Boy activities, or the influence of major seasonal events in the Chinese calendar, such as Spring Festival.

Data reported on the risk and health-seeking behaviours of visitors to the Xiu Boy website must also take into account that, since these data come from a convenience sample, they are not representative either of the wider population of MSM or of all MSM who visited the website. It is also possible that website visitors who declined to complete the risk assessment tool systematically differed from those who completed the assessment in key variables. It should also be noted that this data represents self-reported behavioural data; MSM who completed the assessment may have been motivated by reason of social desirability bias to report safer levels of behaviour than they actually practice. However, participant self-report is a widely used methodology in behavioural research, and studies have suggested that instruments such as Internet-based surveys which do not feature face-to-face interaction may reduce the influence of social desirability bias(Kreuter, Presser, & Tourangeau, 2008). Further, self-reported levels of sexual risk behaviour were broadly similar to behaviours reported in Zhang (2011) and Zu (2013) with the exception that a much higher proportion of Xiu Boy visitors reported having engaged in commercial sex (21.5% versus 5.8% and 5.7%, respectively).


Despite content restrictions and the limited reach of some key global services (i.e. Facebook, YouTube and Twitter) inside of China, information and communication technology platforms, including microsites, online games, digital videos and social media represent an important channel for reaching Chinese MSM and can contribute to increased HTC service uptake and case finding. Indeed, ICT strategies which generate service demand and facilitate service delivery are likely to grow in importance as target audiences increasingly shift to online interactions and funding for resource intensive, venue-based strategies becomes limited. Successful online intervention models hold the promise not only of increased coverage, but also of relatively simple scale-up. It is, however, important not to underestimate the level of resources and technical skill required to implement and sustain these interventions and the importance of partnership and collaboration with governments and service providers if promotion is to translate into service uptake. Finally, it is critical that these interventions be planned with robust monitoring and evaluation measures in place, and that existing monitoring and evaluation systems evolve in order to capture the added value of online intervention activities along with more traditional models such as venue-based peer outreach, in order to further develop an evidence base in support of ICT intervention models.


Berg, R., Tikkanen, R., & Ross, M. (2013). Barebacking among men who have sex with men recruited through a Swedish website: associations with sexual activities at last sexual encounter. Euro surveillance, 18(13).

Chew Ng, R. A., Samuel, M. C., Lo, T., Bernstein, K. T., Aynalem, G., Klausner, J. D., et al. (2013). Sex, Drugs (Methamphetamines), and the Internet: Increasing Syphilis Among Men Who Have Sex With Men in California, 2004-2008. American Journal of Public Health(0), e1-e7.

China Ministry of Health (Ed.) (2012). Beijing.

Gordon, J. (2013). See, Say, Feel, Do – Social Media Metrics That Matter.

Grov, C. (2012). HIV risk and substance use in men who have sex with men surveyed in bathhouses, bars/clubs, and on Craigslist. org: Venue of recruitment matters. AIDS and Behaviour, 16(4), 807-817.

Hull, M. W., Wu, Z., & Montaner, J. S. (2012). Optimizing the engagement of care cascade: a critical step to maximize the impact of HIV treatment as prevention. Current Opinion in HIV and AIDS, 7(6), 579-586.

Jani, I. V., Sitoe, N. d. E., Alfai, E. R., Chongo, P. L., Quevedo, J. I., Rocha, B. M., et al. (2011). Effect of point-of-care CD4 cell count tests on retention of patients and rates of antiretroviral therapy initiation in primary health clinics: an observational cohort study. The Lancet, 378(9802), 1572-1579.

Kilmarx, P. H., & Mutasa-Apollo, T. (2013). Patching a leaky pipe: the cascade of HIV care. Current Opinion in HIV and AIDS, 8(1), 59-64.

Kreuter, F., Presser, S., & Tourangeau, R. (2008). Social Desirability Bias in CATI, IVR, and Web Surveys The Effects of Mode and Question Sensitivity. Public Opinion Quarterly, 72(5), 847-865.

Lau, J. T., Kim, J. H., Lau, M., & Tsui, H. Y. (2003). Prevalence and risk behaviours of Chinese men who seek same-sex partners via the internet in Hong Kong. AIDS Education and Prevention, 15(6), 516-528.

Li, Q., Liu, Y., Zhou, Z., Li, S., Luo, F., Li, D., et al. (2012). Online sex-seeking behaviours among men who have sex with men: implications for investigation and intervention. AIDS and Behaviour, 16(6), 1690-1698.

Lim, S. H., Guadamuz, T. E., Wei, C., Chan, R., & Koe, S. (2012). Factors associated with unprotected receptive anal intercourse with internal ejaculation among men who have sex with men in a large internet sample from Asia. AIDS and Behaviour, 16(7), 1979-1987.

Mtapuri-Zinyowera, S., Chideme, M., Mangwanya, D., Mugurungi, O., Gudukeya, S., Hatzold, K., et al. (2010). Evaluation of the PIMA point-of-care CD4 analyzer in VCT clinics in Zimbabwe. JAIDS Journal of Acquired Immune Deficiency Syndromes, 55(1), 1-7.

Parsons, J. T., Vial, A. C., Starks, T. J., & Golub, S. A. (2013). Recruiting drug using men who have sex with men in behavioural intervention trials: a comparison of internet and field-based strategies. AIDS and Behaviour, 17(2), 688-699.

Population Services International (China) (2010). Routine Behavioural Tracking (RBT) Survey: Draft Summary Report of Key Findings.

Rosser, B. S., Miner, M. H., Bockting, W. O., Ross, M. W., Konstan, J., Gurak, L., et al. (2009). HIV risk and the internet: results of the Men’s INTernet Sex (MINTS) Study. AIDS and Behaviour, 13(4), 746-756.

Rosser, B. S., Oakes, J. M., Horvath, K. J., Konstan, J. A., Danilenko, G. P., & Peterson, J. L. (2009). HIV sexual risk behaviour by men who use the Internet to seek sex with men: results of the Men’s INTernet Sex Study-II (MINTS-II). AIDS and Behaviour, 13(3), 488-498.

Saxton, P., Dickson, N., & Hughes, A. (2013). Who is Omitted from Repeated Offline HIV Behavioural Surveillance Among MSM? Implications for Interpreting Trends. AIDS and Behaviour, 1-12.

Styer, L. M., Sullivan, T. J., & Parker, M. M. (2011). Evaluation of an alternative supplemental testing strategy for HIV diagnosis by retrospective analysis of clinical HIV testing data. Journal of Clinical Virology, 52, Supplement 1(0), S35-S40.

Sullivan, P. S., Carballo-Dieguez, A., Coates, T., Goodreau, S. M., McGowan, I., Sanders, E. J., et al. (2012). Successes and challenges of HIV prevention in men who have sex with men. The Lancet, 380(9839), 388-399.

Suthar, A. B., Ford, N., Bachanas, P. J., Wong, V. J., Rajan, J. S., Saltzman, A. K., et al. (2013). Towards universal voluntary HIV testing and counselling: a systematic review and meta-analysis of community-based approaches. PLoS medicine, 10(8), e1001496.

UNAIDS (2012). Global Report: UNAIDS Report on the Global AIDS Epidemic.

USAID/Health Policy Initiative (2009). Community Perceptions of the Risks and Benefits of Seeking Voluntary Counseling and Testing.

White, J. M., Mimiaga, M. J., Reisner, S. L., & Mayer, K. H. (2013). HIV sexual risk behaviour among black men who meet other men on the internet for sex. Journal of Urban Health, 1-18.

WHO (2012). Service delivery approaches to HIV testing and counseling (HTC): A strategic policy framework.

Wu, Z., Xu, J., Liu, E., Mao, Y., Xiao, Y., Sun, X., et al. (2013). HIV and syphilis prevalence among men who have sex with men: a cross-sectional survey of 61 cities in China. Clinical infectious diseases, 57(2), 298-309.

Young, S. D., Szekeres, G., & Coates, T. (2013). The Relationship between Online Social Networking and Sexual Risk Behaviours among Men Who Have Sex with Men (MSM). PloS one, 8(5), e62271.

Yu, H.-f., An, X.-j., & Tong, J.-y. (2009). Survey of constraints and obstacles for seeking for and receiving VCT services among high risk population in Yunnan province. Chinese Journal of AIDS & STD, 4, 015.

Zhang, D., Bi, P., Lv, F., Tang, H., Zhang, J., & Hiller, J. E. (2007). Internet use and risk behaviours: an online survey of visitors to three gay websites in China. Sexually Transmitted Infections, 83(7), 571-576.

Biographical Statements

Matt Avery is a Regional Technical Advisor on Strategic Behavioural Communications with the FHI 360 Asia-Pacific Regional Office in Bangkok, Thailand. He has been working in HIV prevention with MSM and other key affected populations for more than 10 years.


Gang (Roger) Meng is a head of a Lingnan Huoban a CBO working with MSM and the LGBT community in China. By implementing Internet-based strategies, Lingnan Huoban provided 21,038 HIV tests to MSM during 2008-2013 in Guangdong Province. The CBO was awarded an “Advanced Group” by the Ministry of Health of China in 2013, which is the only grassroots organisations among all of the 156 award-winners.

Stephen Mills, PhD, MPH is Technical Director, Health, Population, and Nutrition, with the FHI 360 Asia-Pacific Regional Office in Bangkok, Thailand. He has been working in HIV programming, capacity building and surveillance for over 20 years.

Matt Avery, Gang Meng & Stephen Mills
Published Online: July 17, 2014
Full Text: HTML, P

User Comments
Author: Two internet-based approaches to promoting HIV counselling and testing for MSM in China - Digital Culture & Education
15 November 2014 07:11:22 AM

[…] Online: July 17, 2014 Full Text: HTML, […]

Author: Sexually Transmitted Diseases: The Hidden Epidemic | Herpes Survival Kit
2 December 2014 06:12:00 AM

[…] transmitted infections (STI) and venereal diseases (VD), are illnesses that have a signific… Sexually transmitted diseases (STD), also referred to as sexually transmitted infections (STI) and v…. WordPress › Error html { background: #f1f1f1; } body { background: #fff; color: #444; […]

Digital Culture & Education (DCE) is an international inter-disciplinary peer-reviewed journal dedicated to the exploration of digital technology’s impacts on identity, education, art, society, culture and narrative within social, political, economic, cultural and historical contexts.

We are interested in empirical and conceptual approaches to theorising globalisation, development, sustainability, wellbeing, subjectivities, networks, new media, gaming, multimodality, literacies and related issues and their implications for how we educate and why. We encourage submissions in a variety of modes and invite guest editors to propose special editions.

DCE is an online, open access journal. It does not charge for article submission or for publication. All manuscripts submitted to DCE are double blind reviewed. Articles are published through a Creative Commons (CC) License and made available for viewing and download on a bespoke page at


Follow us on Twitter at @DigitalCultureE

The scale and speed at which digital culture has entered all aspects of our lives is unprecedented. We publish articles and digital works including eBooks (published under Creative Commons Licenses) that address the use of digital (and other) technologies and how they are taken up across diverse institutional and non-institutional contexts. Scholarly reviews of books, conferences, exhibits, games, software and hardware are also encouraged.

All manuscripts submitted to Digital Culture & Education (DCE) are double-blind reviewed where the identity of the reviewers and the authors are not disclosed to either party.

Digital Culture & Education (DCE) does not have article submission charges. Read more

Manuscripts should include:
1. Cover sheet with author(s) contact details and brief biographical statement(s).

Instructions for Authors

Manuscripts submitted should be original, not under review by any other publication and not published elsewhere.
The expected word count for submissions to the journal is approximately 7500 words, excluding references. Each paper should be accompanied by an abstract of up to 200 words.  Authors planning to submit manuscripts significantly longer than 7500 words should first contact the Editor at

All pages should be numbered. Footnotes to the text should be avoided and endnotes should be used instead. Sponsorship of research reported (e.g. by research councils, government departments and agencies, etc.) should be declared.

Read more

Digital Culture & Education (DCE) invites submissions on any aspect of digital culture and education.  We welcome submissions of articles and digital works that address the use of digital (and other) technologies and how they are taken up across diverse institutional and non-institutional contexts. For further inquiries and submission of work, send an email to editor@