Article Text

Download PDFPDF

Original article
Untested and undiagnosed: barriers to HIV testing among men who have sex with men, Beijing, China
  1. Xuefeng Li1,
  2. Hongyan Lu2,
  3. H F Raymond3,
  4. Yanming Sun2,
  5. Yujiang Jia4,5,
  6. Xiong He2,
  7. Song Fan1,
  8. Yiming Shao1,
  9. Willi McFarland3,4,
  10. Yan Xiao1,
  11. Yuhua Ruan1
  1. 1State Key Laboratory for Infectious Disease Prevention and Control, and National Center for AIDS/STD Control and Prevention (NCAIDS), Chinese Center for Disease Control and Prevention (China CDC), Beijing, P. R. China
  2. 2Beijing Center for Disease Control and Prevention, Beijing, P. R. China
  3. 3San Francisco Department of Public Health, San Francisco, California, USA
  4. 4Institute for Global Health, Vanderbilt University School of Medicine, Nashville, USA
  5. 5Department of Pediatrics, Vanderbilt University School of Medicine, Nashville, USA
  1. Correspondence to Dr Willi McFarland, San Francisco Department of Public Health, 25 Van Ness Avenue, Suite 500, San Francisco, CA 94102-6033, USA; willi_mcfarland{at}


Objectives Undiagnosed HIV presents great potential for the spread of infection. The authors identify the prevalence and correlates of never testing and being unaware of HIV infection in Beijing men who have sex with men (MSM).

Methods Cross-sectional biological and behavioural survey using respondent-driven sampling; 500 MSM were included.

Results HIV prevalence was 7.2% with 86.1% unaware of their infection; 33.2% had never tested. Never testing was associated with lower educational (adjusted odds ratio (AOR) 1.6, 95% CI (CI) 1.1 to 2.5), living in Beijing for ≤3 years (AOR 1.5, 95% CI 1.0 to 2.3), unprotected anal intercourse with most recent male partner (AOR 1.6, 95% CI 1.0 to 2.4), being unaware of the most recent male partner's HIV status (AOR 3.6, 95% CI 2.1 to 6.1) and holding stigmatised attitudes towards persons with HIV (AOR 1.1 per scale point, 95% CI 1.0 to 1.1). Predictors of having undiagnosed HIV infection were being married (AOR 2.4, 95% CI 1.0 to 5.4), living in Beijing for ≤3 years (AOR 3.6, 95% CI 1.5 to 8.4), being unaware of the most recent male partner's HIV status (AOR 6.8, 95% CI 0.9 to 51.6) and holding negative attitudes towards safe sex (AOR 1.1 per scale point, 95% CI 1.0 to 1.1).

Conclusions Recent attention has focused on HIV prevention interventions that depend upon knowing one's serostatus, including viral load suppression, prevention with positives, pre-exposure prophylaxis and seroadaptation. Until the low level of testing and resulting high level of undiagnosed HIV infection are addressed, these tools are not likely to be effective for MSM in China.

  • HIV
  • gay men
  • men who have sex with men
  • China
  • HIV testing
  • undiagnosed HIV infection
  • epidemiology
  • street youth
  • homosexual
  • sexual behaviour
  • risk behaviours
  • AIDS
  • HSV-2
  • epidemiology (general)
  • public health
  • STD control
  • trials
  • STD
  • HCV
  • risk factors
  • viral
  • HBV

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


Recent studies have focused much attention on HIV prevention paradigms that require a very high level and frequency of HIV testing in populations at risk: that viral load suppression through antiretroviral therapy (ART) can reduce transmission at the partnership level1 and incidence at the population level (‘test and treat’),2–4 that HIV-positive persons can change their behaviour to prevent onward transmission after diagnosis (‘prevention with positives’),5 that people can reduce their risk for acquiring or transmitting HIV infection through seroconcordant partner selection or modifying sexual practices when they know their own and their partners' serostatus (‘serosorting’ or ‘seroadaptation’)6 7 and that taking ART by HIV-negative persons can prevent acquisition of infection (‘pre-exposure prophylaxis’ or ‘PrEP’).8 All these prevention approaches require persons know their HIV serostatus in order for them to be effective. If levels of HIV testing and serostatus awareness are low in a population, then these tools are not likely to change the course of the epidemic and prevention approaches that do not require knowledge or disclosure of serostatus (eg, consistent condom use) would have greater traction.

Men who have sex with men (MSM) in Asia, including China, urgently need appropriate and effective HIV prevention tools. HIV prevalence is high and rising among MSM throughout the continent.9 One poignant example is the astonishing situation among MSM in Bangkok, with HIV prevalence rising from 17.3% in 2003 to 30.8% in 2007.10 High prevalence and direct measures of incidence demonstrate that HIV transmission continues rapidly among MSM in China.11–14 Between 2007 and 2009, China's projections have increased the attributed proportion of the epidemic among MSM from 12.7% to 32.5%, suggesting that MSM are becoming the predominant group affected by HIV in the country.15 Meanwhile, recent HIV testing (in the past 12 months) coverage among MSM in China remains low, with figures typically under 30%.16

The new HIV treatment paradigm also increasingly embraces early initiation of ART to maximise the long-term health benefits to persons living with HIV.17 Both public health prevention and patient care may be aligned in prioritising the earliest possible diagnosis of HIV infection. We therefore conducted a behavioural and serological community-based survey of MSM in Beijing, China, in order to determine who remains untested and who remains undiagnosed among persons HIV infected. We further probed the possible reasons for why MSM avoid HIV testing and garnered MSM opinions on how to increase testing uptake. In doing so, our overarching aim was to guide interventions to tackle what may be the major barrier to HIV prevention and care for MSM in China.


Study subjects and recruitment

We conducted a cross-sectional survey using respondent-driven sampling (RDS) among MSM with interviews conducted at a voluntary counselling and testing site run by the Beijing Center for Disease Control between September and October 2009. Participants were eligible if they self-reported as being 18 years of age or older, a Beijing resident, having had sex with a male in the prior 12 months, having a valid recruitment coupon and were able to provide informed consent. The methods of RDS have been previously described18 and applied to multiple surveys of MSM in China and throughout the world.19 20 Enrolment began with choosing MSM seeds to initiate long chains of recruitment by referring eligible acquaintances. Seeds were chosen based on the diversity of social networks and demographic characteristics, with candidates articulate and supportive of the study aims. We selected seven seeds from bars, parks, bathhouses and over the internet—venues and websites frequented by MSM. These MSM in turn referred three others among their social networks, and so on until 500 participants enrolled and equilibrium was achieved (ie, when the composition of the sample does not change with additional waves of recruitment). For the present study, we tracked age, education, employment, residence, marital status, sexual orientation, multiple partners and unprotected anal sex to equilibrium. Incentives included 30 RMB for completion of the survey (primary incentive) and 20 RMB for each of up to three eligible MSM they were able to recruit (secondary incentive). HIV counselling and testing was offered to participants after completion of each survey questionnaire with results available within 1 to 3 weeks by face-to-face consultation or by telephone according to their preference. Appropriate risk-reduction counselling was provided to all participants, and referrals for appropriate medical, mental health and social support services were provided to persons as needed.

The protocol was reviewed and approved by the human subjects ethical committees of the National Center for AIDS/STD Control and Prevention (NCAIDS) in China and Vanderbilt University and the University of California San Francisco in the USA.


Participants completed a self-administered, computerised structured questionnaire based on previously used instruments21–23 and collected information about respondents' demographic characteristics, HIV risk behaviours, HIV testing history and reasons for seeking or avoiding HIV testing. Men were asked if they disclosed their serostatus to partners, and results are reported according to their responses regardless of whether they had ever tested or tested differently from their self-reported serostatus in our study. Two psychosocial scales were adapted and translated from instruments validated among MSM in Africa and North America.22 23 One scale was on attitudes and perceptions of abilities for safe sex and was measured by 15 items asking to what extent MSM were confident in taking steps towards safer sex using a 4-point Likert scale (strongly disagree, somewhat disagree, somewhat agree and strongly agree). For example, “I find it difficult to have safer sex when high or drunk.” The internal reliability was high (Cronbach's α 0.92) in our study. In our analysis, the scale was inverted towards increasing score being negative attitudes and perceptions of abilities towards safe sex. The second scale was on negative attitudes towards persons with HIV/AIDS and included questions such as ‘People with AIDS should be isolated from other people’ (22 items, Cronbach's α 0.82 in our study).

Laboratory methods

After completion of the questionnaire, blood was drawn for HIV testing. All specimens were tested for HIV antibody by ELISA (Vironostika HIV Uni-Form plus O; bioMerieux, Mercy L'Etoile, France) and western blot (HIV Blot 2.2 WBTM; Genelabs Diagnostics, Singapore). MSM were considered HIV infected if both tests were positive.

Statistical methods

The framework of our analysis was to characterise MSM according to two key outcomes: those who have never previously tested for HIV and HIV-infected MSM who were unaware of their HIV-positive serostatus (ie, undiagnosed) prior to our survey. In both analyses, we conducted bivariate and multiple logistic regression analyses using SAS V.9.1 and Stata V.11.0 software. In multiple logistic regression analyses, we included variables that were significant in bivariate analysis (p<0.05), were considered a priori important characteristics of testing or potential confounders. Because our analysis centred on assessing internal relationships rather than population point estimates and also incorporates complex scales, analysis was conducted outside of RDSAT (


Sample description

Most MSM participating in our survey were over the age of 25, had high school education or more, were unmarried, employed, lived in Beijing for >3 years and had homosexual sexual orientation (table 1). Unprotected anal intercourse (UAI) with their most recent male partner was reported by 35.4%, UAI with a casual partner was 24.0% and 15.0% had unprotected vaginal sex with their most recent female partner. A minority (26.4%) reported that they knew the HIV serostatus of their most recent male sexual partner; more (48.4%) reported disclosing their HIV serostatus to their most recent male partner. Nearly all MSM (94.2%) knew where to get an HIV test, yet nearly one-third (33.2%) had never tested. Among all participants, 13.8% tested within the past 3 months, 28.4% between 3 and 6 months and 11.0% between 6 and 12 months. The prevalence of HIV in our sample was 7.2% according to serological tests performed for this study.

Table 1

Sample characteristics, men who have sex with men, Beijing, China, 2009 (N=500)

Factors associated with never being tested

Table 2 shows correlates of never having tested for HIV. In multivariate analysis, independent predictors of never testing were low educational level (adjusted odds ratio (AOR) 1.6, 95% CI 1.1 to 2.5), living in Beijing for ≤3 years (AOR 1.5, 95% CI 1.0 to 2.3), having UAI with their most recent male partner (AOR 1.6, 95% CI 1.0 to 2.4), being unaware of their most recent male partner's HIV status (AOR 3.6, 95% CI 2.1 to 6.1) and holding stigmatised attitudes towards persons living with HIV (AOR=1.1 per scale point, 95% CI 1.0 to 1.1).

Table 2

Correlates of never testing for HIV, men who have sex with men, Beijing, China, 2009 (N=500)

Self-reported and actual HIV status

The low level of testing resulted in a high proportion of undiagnosed HIV infection; most persons infected with HIV were unaware of their status. Of the 36 MSM testing positive through our study, only five (13.9%) reported being previously diagnosed (ie, tested before and knew their positive status), 17 (47.2%) perceived themselves to be HIV negative based on a previous test, 11 (30.6%) had never tested and three (8.3%) had tested but did not know their result. Re-examining self-report of disclosing HIV status among the HIV infected but undiagnosed participants suggests that seven HIV-positive MSM may have disclosed the wrong serostatus to partners, although the timing of infection is uncertain. Of the 464 men who tested HIV negative in our study, 303 (65.3%) reported having had a previous HIV-negative test and 161 (34.7%) reported not knowing their serostatus. Of these, 155 never tested before and six tested previously but did not get their results.

Factors associated with undiagnosed HIV infection

Table 3 presents factors associated with undiagnosed HIV infection. Our perspective for this model was to characterise persons most likely to be infected among those persons we wish to reach with testing (ie, if they already know they are HIV positive, then they do not need to test). Under this rationale, the model excludes the five men who knew their HIV-positive status. MSM who were married to a woman (AOR 2.4, 95% CI 1.0 to 5.4), lived in Beijing for ≤3 years (AOR 3.6, 95% CI 1.5 to 8.4), were unaware of the most recent male partner's HIV status (AOR 6.8, 95% CI 0.9 to 51.6) and held negative attitudes towards safe sex were more likely to have undiagnosed HIV infection (AOR 1.1 per scale point, 95% CI% 1.0 to 1.1).

Table 3

Independent predictors of undiagnosed HIV infection, men who have sex with men, Beijing, China, 2009 (undiagnosed HIV positive vs HIV negative, n=495)

Barriers to HIV testing

Our participants identified multiple barriers as well as potential facilitators of HIV testing among MSM in Beijing (table 4). Among MSM with undiagnosed HIV infection, the three most common barriers to testing mentioned were fear of knowing they might be HIV positive, unsure where to go to get tested and fear of discrimination if positive. Two of these factors relate in part to perceived negative attitudes towards persons with HIV. Four other barriers were also related to fear of stigmatisation or discrimination (eg, afraid of being seen by friends at a testing site, not wanting to go to a site, ie, designated specifically for HIV testing, worried people will think they have AIDS if they know they tested and fear that results will not be held confidential). Nearly two-thirds of those with undiagnosed HIV infection said MSM do not test because they do not consider themselves at risk for infection. A majority felt MSM do not test because they perceive they cannot afford treatment if positive. HIV-negative MSM, whether previously tested or untested, largely concurred with these barriers to testing.

Table 4

Barriers and promoting factors for seeking HIV testing, men who have sex with men, Beijing, China, 2009 (n=495)

Facilitators of HIV testing

When asked their opinions on what would make more MSM test, nearly all endorsed ensuring confidentiality or allowing that tests can be done anonymously (factors addressing perceived stigma), providing free or low-cost testing and increasing people's knowledge of HIV/AIDS (table 4). Several other potential test-promoting factors were related to making testing more routine or universal (in part addressing perceived stigma), including promoting the test a standard part of medical practice, if many other people were to test or if it were legally required. Two other factors were directly related to the need to reduce stigma towards persons with HIV, including more sympathetic health personnel, less community discrimination or more personal connection to someone who has AIDS as test motivating factors.


According to our data, several popular prevention paradigms would not operate effectively for MSM in Beijing, given the low levels of testing and high levels of undiagnosed infection. One in three MSM had never tested for HIV. For comparison, the figure is more than three times the level measured among MSM in 21 cities of the USA in 2008 (10%).24 Compounding the negative impact of low testing uptake on the epidemic, men never testing were also more likely to engage in unprotected anal sex and have unknown serostatus partners. Stemming from low levels of testing, more than four-fifths of HIV-infected MSM in our survey were unaware of their infection. This compares with 44% of HIV-infected MSM in the USA being unaware and was higher than witnessed in Baltimore (73%) where HIV prevalence among MSM was the highest in that country.25 Further compounding the problem, men with undiagnosed with HIV infection were also unaware of their most recent partner's serostatus and held more negative attitudes towards safe sex practices. Massive efforts will be needed to encourage testing at all and frequent testing among MSM to reach the levels of being undiagnosed comparable to San Francisco (19%),24 where evidence suggests that viral load suppression due to treatment and serosorting appear to be taking effect on reducing HIV incidence.3 7

On the positive side, we were able to identify several barriers to testing among MSM that may be amenable to interventions. First, our study suggests that several structural changes in the mode of delivery are likely to improve testing among MSM in Beijing. Regardless of prior testing history, current serostatus and undiagnosed infection status, MSM agreed that providing free testing, anonymous options and routine integration into other health services would reduce barriers and stimulate greater uptake. Second, our findings identified priority groups for targeted promotional messages and outreach testing, including recent migrants to the city, those with lower level of education and married MSM. A third and challenging factor is the stigmatisation of persons living with HIV. Structural changes in the modes of delivering HIV testing may also help to address the barrier of stigmatisation (ie, availability of anonymous testing and the incorporation of testing into routine healthcare) for MSM to seek and get tested in the short term. However, changing societal attitudes for the long term, a difficult prospect, will be needed to link and retain HIV-positive persons in specialised care.

We recognise limitations of our study. Because HIV testing was part of the study design, participation may be biased by the same factors that prevent some MSM from seeking testing. If persons less likely to test are also less likely to participate in our survey, then the true levels of HIV testing in the population at large may actually be lower and the proportion of undiagnosed infection even higher than reported here. However, an opposite effect may also be possible if MSM who already know their HIV positive are less likely to participate in the survey. Another potential bias may stem from the monetary incentive having relatively stronger draw among persons of lower socioeconomic status, as noted in other RDS surveys.20 Since our data found lower education level associated with never testing, the bias resulting from preferential enrolment of lower income men is likely to overstate the levels in the population as a whole. While the net effect of these competing biases would be exceedingly difficult to quantify, we believe that our findings present a conservative estimate of undiagnosed infection, as similar and even lower levels of HIV testing have been found among MSM in other cities in China.16 21 26 Finally, we recognise the limitations of RDS and other methods used to approximate representative samples of hidden populations. RDS is not a simple random sample, but rather subject to assumptions that are difficult to verify, and final results do not have a gold standard to which to compare.

We believe that such potential biases do not obscure the evidence of large numbers of MSM who have never tested and HIV-infected MSM who remain undiagnosed, in China. Our study highlights the fact that HIV testing stands at the gateway to a comprehensive plan of prevention and care. We have characterised a rate-limiting step in controlling the HIV epidemic among MSM in China, who may currently be its predominant risk population. While increased HIV testing is clearly a priority, it is important to remember that serological diagnosis is not HIV prevention in and of itself. Testing needs to be coupled with appropriate counselling and linkage to other health, prevention and social welfare programmes. Persons diagnosed as having HIV need to be provided affordable treatment delivered in a non-stigmatising environment. The current expanding stage of the HIV epidemic among MSM in China may be comparable to that experienced by their counterparts in the West in the early 1980s, but with the important difference that prevention tools not known 30 years ago can be brought to bear. These include the more recent interventions that entail ART and serostatus awareness1–8 and those that can increase condom use while one's own and one's partners' serostatus remain unknown.27

Key messages

  • Recent biomedical and behavioural breakthroughs in HIV prevention may not benefit many men who have sex with men in China, given the low level of HIV testing and serostatus awareness.

  • HIV testing that is free, anonymous and integrated into routine healthcare services may substantially increase its uptake among men who have sex with men in China.

  • Special efforts will be needed to provide HIV testing to men who have sex with men who are migrants within China, poor and married.



  • Funding This study was funded by grant R01 AI078933 from the National Institutes of Health of the USA and by the Ministry of Science and Technology of China (2008ZX10001-004, 2008ZX10001-010 and 2009DFB30420).

  • Competing interests None.

  • Patient consent Obtained.

  • Ethics approval National Center for AIDS/STD Control and Prevention (NCAIDS) in China and Vanderbilt University and the University of California San Francisco in the USA.

  • Provenance and peer review Not commissioned; externally peer reviewed.