HIV testing frequency and associated factors among five key populations in ten cities of China: a cross-sectional study

As of October 2019, China’s National Health Commission has reported an estimated 958,000 people living with HIV who have been tested and thus know their sero-status [1]. Increasing HIV testing among at-risk individuals and linking HIV-positive individuals to care has been identified as a critical strategy to mitigate the ongoing transmission of HIV [2, 3]. The WHO recommends that key populations, including men who have sex with men (MSM), persons who exchange sex for money, and drug users, should be tested for HIV at least once a year [4].

Regular HIV testing is the gateway to detecting people living with HIV promptly, but more research is needed to learn about the risk-profile and characteristics of frequent testers. Studies on the facilitators and barriers to HIV testing among different key populations have found links with socio-demographic factors (e.g., age, education level, marital status) as well as behavioral factors (e.g., multiple sexual partners, unprotected sex, drug use)[5‐8]. However, there is a dearth of literature about factors associated with frequent HIV testing among key populations other than MSM. In and of itself, high-risk sexual behavior is considered a key correlate of frequent HIV testing, yet its effect is inconsistent [9‐12]. For instance, studies in London and Germany have found that frequent HIV testing is associated with reported unprotected anal sex or condomless sex with casual partners [9, 11]. But a study in Peru indicated that those reporting unprotected anal sex were less likely to test frequently [10].

In China, the HIV epidemic is mainly concentrated among MSM, female sex workers (FSWs), people who use drugs (PWUD) and male patients of sexually transmitted disease clinics [13]. These are considered high-risk groups and are included in Chinese HIV Sentinel Surveillance System [14]. In addition, data shows that about 25% of heterosexual HIV transmission is through sexual contact with one’s partner, indicating the transmission between sero-discordant couples plays a non-trivial role of new infections [15]. Hence, one of the key populations included in this study is sero-negative partners among sero-discordant couples, or SNPs.

Current policies in China focus on expanding coverage of testing without offering specific guidelines on recommended HIV testing frequency. This is potentially problematic because previous studies have suggested HIV testing uptake is insufficient among certain key populations such as PWUD (25.8% in 2018) [8] and male clients of FSWs (23.8% in 2019) [16]. With respect to MSM, a study in Beijing showed 71% reported ever testing for HIV and 52% reported ≥ 2 HIV testing in their lifetimes [17]. Another study of MSM in Guangzhou defining frequent testers as those who did so at least twice per year found that 44.3% were tested frequently [18]. Some of psychosocial factors associated with frequent testing were having disclosed sexual orientation to non-gay friends and having lower internalized homophobia [18].

There have been few studies exploring the socio-behavioral factors associated with HIV testing frequency, especially in China. Again, the literature focuses primarily on MSM, ignoring other key populations at risk of HIV acquisition. It remains unclear whether the socio-behavioral factors impacting HIV testing frequency are consistent among these other key populations. This study aims to fill that research gap by examining and comparing HIV testing frequency and associated factors among five key populations in China that includes not only MSM, but also FSWs, PWUD, SNPs, and men who have casual sex with women (MCSW). Findings from this article could help public officials improve practice and potentially re-allocate resources to decrease the risk of further HIV transmission among these high-risk subgroups.

The purposes of this study was to explore HIV testing frequencies amongst five key populations in ten cities of China. Specifically, we explored the socio-demographic and behavioral factors associated with HIV testing frequencies using PPOMs from two perspectives, one considering whether individuals had taken an HIV test in the past 12 months, and one considering whether individuals had taken two or more HIV tests, which we refer to as being a frequent HIV tester.

The percentages of individuals reporting to have taken an HIV testing in the past 12 months differed amongst the five key populations. As compared with MSM, FSWs and SNPs were more likely to have taken an HIV test, while MCSW were less likely. Interestingly, again relative to MSM, we found that FSWs were less likely to be frequent HIV testers even though they were more likely to have taken an HIV test. The HIV testing uptake rate for MSM in the past 12 months was 69.7% in our study, which is significantly higher than the 38% found in a 2012 meta-analysis [25]. This may be attributable to the Chinese government strengthening its public health efforts in promoting HIV testing in recent years. SNPs were the most proactive group of HIV testers, which is possibly because sero-positive people are obligated to disclose to their spouses according to Chinese Regulation on the Prevention and Treatment of HIV/ AIDS. A study in Yunnan province of China found 76.32% of SNPs took an HIV test within 3 months after their spouse confirmed an HIV infection [26]. For MCSW, the proportion of HIV testing was lower than others at 34%. Similarly, a study in 2019 reflected the low uptake of HIV testing (23.8%) among male clients of FSWs in China [16]. This may be due to commercial sex being illegal in China [27] and the hard-to-locate nature of this population. Future interventions and HIV testing promotion efforts should focus on this population. For FSWs, 76.8% reported testing for HIV in the past 12 months in our study, which was higher than the 48% found in a study by Chow in 2015 [28]. At the same time, only 11.7% of FSWs in our study could be characterized as frequent testers. Complex psychosocial and demographic characteristics may affect their testing behavior, some of which are interpreted in detail below.

Stratified analyses showed that certain socio-demographic factors inconsistently influenced HIV testing frequency among the different populations. For MSM and SNPs, those with higher education were more likely to have taken an HIV test, which is consistent with previous studies [29, 30]. People with higher education are more likely to have access to resources and thus more likely to see the benefits of knowing their HIV status. Conversely, we found that FSWs with higher education were less likely to have taken an HIV test, which is consistent with a study in Southwest China [31]. A possible explanation was that “active testing” conducted by local CDC offices had targeted certain commercial sex venues, such as those that provide more “explicit” commercial sex services (e.g., massage parlors and streets). In other venues such as hotels and night clubs, where commercial sex is hidden under the cover of other services, the outreach for “active testing” is less common. And the literature suggests that the FSW with less education are more likely to work in venues that offer more “explicit” commercial sex service [32].

The impact of monthly income on HIV testing frequency varied for FSWs and PWUD. We found FSWs who earned ≤ 3000 CNY monthly to be more likely to take at least one HIV test but to be less likely to be frequent HIV testers compared with their counterparts earning > 5000 CNY. Low-income FSWs are more likely to report high rate of condomless sex with their clients (on occasion, sacrificing condom use for extra payment) and are known to suffer severe stigmatization and marginalization in society [33]. More frequent high-risk sexual experiences may compel them to test for HIV but limited resources, due to poverty, may prevent them to being frequent HIV testers [34]. PWUD with higher income were more likely to take at least one HIV test and to be considered frequent testers. It is possible that more disposable income enables them to more easily access illicit drugs and report high-risk sexual behavior after using them [35].

With respect to age, we found that elder PWUD were less likely to have reported an HIV test in the past 12 months, which is consistent with a study in Shanghai [36]. Younger people tend to engage more frequently in high-risk activities but also have a greater chance of accessing HIV-related information. However, our study found HIV testing rates to be higher amongst elder FSWs. Working in the sex industry for a longer period and starting sex work at older ages were reported to be associated with higher HIV testing uptake [34], which suggests that the willingness to take an HIV test may increase with time and correlate closely with HIV risk perception among FSWs.

Evidence shows that heavy alcohol use can impair judgment and directly affect the brain resulting in reduced inhibitions and diminished risk perception [37]. In addition, alcohol consumption in entertainment venues such as karaoke bars, dancing hall, nightclubs, is sometimes accompanied by illicit drug use and unwanted sexual behaviors [38], which increases the likelihood of HIV transmission. In our study, associations with alcohol use were mixed. We found more frequent alcohol consumption appeared to facilitate HIV testing among FSWs and PWUD, but actually decreased the rate of HIV testing among MCSW.

In this study, high-risk sexual behaviors among certain key populations were found to be linked to more frequent HIV testing, which is consistent with studies in London and Germany [9, 11]. Specifically, we found that MSM who had condomless anal sex with men in the past 6 months, FSWs who had ≥ 7 male clients during a week, and PWUD who had condomless sex with multiple partners after using drugs in the past year were all more likely to be frequent HIV testers. Of course, from a public health perspective, the desired goal is to reduce high-risk behaviors, not strictly to increase the frequency of HIV testing. So individuals pursuing high-risk behaviors would surely benefit from other interventions such as public relations campaigns on the importance of condom use preventing HIV transmission as well as the potential benefits of biochemical approaches such as non-occupational post-exposure prophylaxis (nPEP) and pre-exposure prophylaxis (PrEP) [39, 40].

Several limitations of this study should be acknowledged. First, self-reported data, some of which is sensitive (e.g., sexual behavior, illicit drug use), may not always be accurately reported. Second, a convenience sampling approach was used to recruit participants, which may impact the HIV testing percentages that would have been reported under a (much less feasible) probability-based sampling approach. Third, we only focused on comparing the socio-demographic and behavioral factors associated with HIV testing frequency in this study without taking structural factors (e.g., stigma) into consideration. Fourth, we are unable to distinguish between HIV self-testing and facility-based testing in this study. It is possible that results and insights could differ if we were instead able to make that distinction for our recruited participants [41].

Rates of HIV testing varied amongst the five key populations considered in our study. Compared with MSM, MCSW were less likely to have reported taking an HIV test within the last 12 months, while SNPs were more likely to be frequent HIV testers and FSWs were more likely to take have taken an HIV test but less likely to be frequent HIV testers. The factors associated with HIV testing frequency were not completely consistent for each population. In order to optimize HIV prevention and care, it is important to identify high-risk subgroups reporting low rates of HIV testing. That said, public health practitioners should also focus on behavioral change for individuals in high-risk subgroups even if data demonstrates they tend to more frequently test for HIV. Intensive and tailored behavioral interventions coupled with biochemical prophylaxis would be helpful in that regard.