High HIV prevalence and the internet as a source of HIV-related service information at a community-based organization in Peru: a cross-sectional study of men who have sex with men

The estimated prevalence of HIV in Latin America in 2014 was 0.2–0.7 % [1] and is particularly concentrated among urban men who have sex with men (MSM) [2]. According to the 2014 UNAIDS progress report, between 7 and 20 % of MSM in Latin America are infected with HIV [3]. Approximately half of all new infections in the region stem from unprotected anal intercourse (UAI) between men [4, 5]. In 2011, the prevalence of HIV among MSM in Peru was thirty times greater than that in the general adult population (12.4 % vs. 0.4 %) and was most prevalent in the capital city of Lima [2, 6].

Diagnosis is the critical first step in getting infected individuals into HIV care, but less than 50 % of MSM in Lima have been tested for HIV [7]. Among those who do present for testing, as few as 59 % attend a follow-up visit to receive their test results [8]. Moreover, less than a quarter of MSM know the HIV status of at least one of their last three partners [9]. Reported reasons for low testing levels include fear of receiving a positive diagnosis, resulting in lifelong dependency on HIV treatment, drastic lifestyle changes, and even death [7, 10].

The Internet is well established as a mechanism for promoting novel HIV prevention interventions targeting MSM in low and middle-income countries (LMIC) due to its accessibility and cost-effectiveness [11‐14]. Peru is classified as an upper-middle income country by the World Bank [15]. According to World Bank data from 2014, 40.2 per 100 people in Peru are internet users, defined as individuals who have used the Internet from any location in the last 12 months via computer, mobile phone, personal digital assistant, games machine, digital TV, etc. [15]. In the last decade, mobile-cellular subscriptions have catapulted from 8.63 to 98 per 100 inhabitants in Peru, per the United Nations International Telecommunication Union [16].

While World Bank data identifies only 40 % of Peruvians as Internet users, studies have reported the utility of the Internet as a tool to access high-risk MSM in Peru as early as 2007 [13, 15]. This study also highlighted that 80 % of these people access the Internet via cabinas publicas, or Internet cafes, which are generally concentrated in the same urban areas where most MSM live in Peru [13]. According to a 2011 study, a high-risk subset of men not reached by traditional prevention methods uses the Internet to seek both sexual partners and information about HIV and other sexually transmitted infections (STI) [12].

Because Internet use by MSM is correlated with high-risk sexual behavior and is increasing rapidly in LMICs, understanding factors that distinguish MSM who regularly use online resources related to sexual health from those less likely to use them would enhance strategic planning for public health interventions [13, 17‐19]. Community based organizations (CBOs) in resource-limited settings need to know how and where to target their outreach activities to reach the highest risk populations and maximize the impact of limited resources. This study aims to describe both the HIV prevalence and risk profiles of MSM who sought HIV-related servicesat Epicentro Salud (Epicentro), a community-based organization in Lima, Peru, comparing HIV prevalence among those who found out about Epicentro through online media and those who reported sex work to those who became aware of the organization through word of mouth.

We found that nearly 23 % of Peruvian MSM who found out about HIV services at Epicentro via online sources were HIV-infected. This was substantially higher than other MSM who received the same information via friends or partners (16 %), a significant difference even after accounting for confounding factors including Internet behaviors (aOR = 1.61; 95 % CI: 1.19–2.18). This population of MSM represents an important group for HIV surveillance and prevention activities. As expected, the observed HIV prevalence in the sex worker group was also higher (19 %), but did not differ substantially from MSM who received the same information via friends or partners.

The overall HIV prevalence of 18.3 % from this study falls within the range previously reported by studies of MSM in Lima (10–30 %) for which participant recruitment varied across studies (e.g., general adult clinics, clinics catering more to MSM, or community-based recruitment) [20‐23]. Of additional interest, most sexual risk profiles between the in-person and online groups were similar, but the online group was less likely to use alcohol and drugs during sex or to exchange money or goods for sex, while more often reporting internet-related behaviors, as expected. These differences, however, did not explain the observed association between finding out about Epicentro via online sources and having a higher HIV prevalence. Thus, MSM who were responsive to online information about Epicentro services may represent a much broader, but still at-risk, group of MSM than those who specifically seek out sexual partners online.

Among MSM who identified as sex workers, the observed HIV prevalence (19.3 %) falls between that of the in-person (15.9 %) and online (23.2 %) groups. Not surprisingly, sex workers reported more frequent sexual risk behaviors, yet they also reported higher use of HIV prevention behaviors, including HIV testing and regular condom use. This group also reported lower prevalence of STI symptoms in the last 6 months. These findings are consistent with recent studies looking at condom use and other prevention strategies employed by MSM sex workers and their clients in Peru and other parts of the developing world [24‐26]. The high frequency of risk prevention techniques reported by sex workers is a clear sign that STI prevention outreach and education to this high-risk group has had an impact, and the lessons learned from this work could potentially be applied to other groups.

Findings from this study specifically suggest that MSM who actively engage online media (for example, to identify MSM or HIV-related services), even for reasons other than finding sexual partners, are at increased risk of HIV infection. Numerous studies looking at the risk profiles of MSM in the United States and other high income countries suggest those who use the Internet to look for sex have an increased risk for HIV infection due to the anonymity provided by this media. However, few studies have sought to corroborate these results in a low-resource setting [17‐19, 27‐30]. The reason why our study found engaging online media for HIV-service information to be related to risk of testing positive for HIV, independent of Internet use to find sexual partners, is not fully clear. It may reflect concerns about remaining anonymous about one’s sexual orientation and HIV risk, which in turn may affect HIV service utilization, rather than specific behaviors that increase risks of HIV infection directly. Some studies have suggested that stigma, discrimination, or disclosure concerns around sexual orientation may negatively impact HIV prevention and services utilization [31‐33].

Our finding that MSM who reported using online media as one of their sources of HIV-related information have a substantially higher prevalence of HIV has important implications for how public, private and not-for-profit agencies strategically plan their HIV prevention and testing service delivery and marketing. Targeting resources via online and social media routes facilitate access to an important risk group with increased need of HIV testing and subsequent treatment or prevention interventions. A small but growing literature has begun to explore the use of web-based HIV interventions via randomized controlled trials (RCT) or quasi-experimental studies among MSM and conducted entirely over the Internet [11]. Most studies are from the United States or other high-income countries and report a reduction in one or more risk behaviors or increases in HIV testing, knowledge, self-efficacy, outcome expectancies, and/or disclosure [11, 34, 35]. One small RCT (n = 187), conducted in Peru, found a 5-min online video successfully moved participants through stages of change for HIV testing (e.g., contemplation to preparation stage) and was associated with both making a testing appointment and receiving an HIV test among non-gay identified MSM (no association among gay-identified MSM) [14]. The advantages of online interventions lie in their ability to engage large numbers of individuals from hard-to-reach populations such as MSM in a consistent and cost-effective manner, while providing the anonymity that is not possible with clinic- or peer-based interventions [11]. The combined effects of increased uptake of modern technology and personalized developments in HIV prevention strategy, including things like the release of the OraQuick^TM in-home HIV test in the United States, should make it easier for MSM reluctant to access services in-person to take advantage of similar services in the privacy of their own homes.

This study has some limitations. First, the study used a cross-sectional design, which limits inference about a causal relationship. Still, the high HIV prevalence among the online group is concerning and indicates the need for ongoing prevention and testing messages geared towards MSM using online social media and other sites. Second, the study data were collected for routine clinical and testing purposes, and consequently did not have sufficient detail to more fully explore Internet use for HIV information or risky sexual behaviors. Third, the Epicentro questionnaire asked for information regarding all sexual risk behaviors and reported Internet use in the past three months, whereas questions referring to sexual role pertained to the past six months. It is unlikely, however, that the difference in these time periods would add any bias to study results. Finally, there is a possibility for misrepresentation inevitable in any study relying on self-report, whether this is due to shame, recall bias, intent to please the counselor conducting the survey, or just from providing quick responses without careful consideration. For example, it is possible that there are some sex workers in the in-person group that did not identify themselves as such. Misclassified sex workers would likely be in the in-person group (reporting having heard about Epicentro through a friend) because all of the sex workers except one originally fell into that group prior to being separated as an individual category. Such misclassification would likely inflate the HIV prevalence of the in-person group, meaning that the observed association between finding out about Epicentro HIV-related services via the Internet and HIV prevalence may be underestimated.

Increasing Internet uptake in low-resource settings where there is high HIV prevalence among MSM represents a unique opportunity for CBOs in this setting to employ low-cost, internet-based HIV prevention and testing messages. In the context of Lima, Peru, education and outreach for HIV/STI prevention and testing has traditionally been conducted through peer education, which requires investment of substantial human and financial resources. The transition of some of this messaging to Internet marketing may permit these CBOs and other public health entities to focus limited resources as well as broaden the number of individuals reached, particularly among high-risk groups such as MSM who would benefit the most. Further research is needed to explore the differential risk profiles for MSM that actively use the Internet as a resource for information about sexual health and to seek sexual encounters so that interventions can be more specifically tailored to the target population. Organizations providing HIV prevention and testing services should also continue to test Internet-based marketing to evaluate both its reception and usability in this unique MSM population.