Abstract
This article examined differences in condom use during anal intercourse among men who have sex with men (MSM) and male-to-female transgender women in Jakarta, Indonesia. A cross-sectional design, structured interviews, and hierarchical linear modeling were used to examine condom use among MSM recruited from entertainment places (EPs; e.g., discotheques/dance clubs/karaoke bars), massage parlors (MPs), and among transgender women who congregated and/or sought sexual partners on streets/parks (S/P). The sample consisted of 91, 97, and 114 of MSM-EP, MSM-MP, and transgender-S/P, respectively. Respondents reported on 641 unique sexual partner encounters, which were “nested” within 302 respondents. Reported condom use was high, 66%, 84%, and 83% for MSM-EP, MSM-MP, and transgender-S/P, respectively, and varied across type of respondent. At the individual level, depressive symptoms and history of physical abuse during childhood and adulthood were associated with lower condom use (p < .05). By contrast, having a higher level of education was associated with more condom use (p < .05). At the partner level, condom use was associated with type of partners and the use of club drugs before sex. HIV-prevention efforts should take into account the multilevel determinants of condom use within these populations.
Keywords
Introduction
Unprotected anal intercourse remains a major HIV-transmission vector among men who have sex with men (MSM) and male-to-female transgenders worldwide (Beyrer et al., 2012; Herbst et al., 2008; Koblin et al., 2003; Koblin et al., 2006; Morineau et al., 2011; Operario, Nemoto, Iwamoto, & Moore, 2011; De Santis, 2009; Zhang et al., 2007). Empirical data on factors associated with unprotected anal intercourse are well documented. These include drug use, particularly club drugs (e.g., methamphetamine and/or ecstasy; Boeri, Sterk, & Elifson, 2004; Diaz, Heckert, & Sanchez, 2005; Lyons, Chandra, Golstein, & Ostrow, 2010; Mackesy-Amiti, Fendrich, & Johnson, 2008; 2010; Ramirez-Valles, Garcia, Campbell, Diaz, & Heckathorn, 2008), alcohol use (McKirnan, Vanable, Ostrow, & Hope, 2001; Stall et al., 2001), the use of sex drugs, such as Viagra (Chu et al., 2003; Fisher, Reynolds, & Napper, 2010), psychological factors such as depression and stigmatization (Chen, Choe, Chen, & Zhang, 2005; Clements-Nolle, Guzam, & Harris, 2008; De Santis, 2009; Diaz, Ayala, Bein, Henne, & Marin., 2001; Neilands, Steward, & Choi, 2008; Nuttbrock et al., 2013; Salomon et al., 2009), and history of forced sex and violence during childhood and adulthood (Guadamuz et al., 2011; R. J. Johnson et al., 2006; Kosenko, 2011; Lombardi, Wilchins, Priesing, & Malouf, 2001; Maman, Campbell, Sweat, & Gielen, 2000; Nuttbrock et al., 2013; Ratner et al., 2003).
A growing literature confirms the importance of partner influences on condom use that may shape and/or protect HIV-risk behavior among individuals. These include partners’ beliefs and behaviors (De Santis, 2012; Eaton, West, Kenny, & Kalichman, 2009); type of sexual partner, with lower use of condoms when sex partners are “nonpaying (e.g., husband, boyfriend or casual)” than “commercial” (Clements-Nolle et al., 2008; De Santis, 2012; Mi, Wu, Zhang, & Zhang, 2007; Nemoto, Operario, Keatley, & Villegas, 2004; Wilson, Garofalo, Harris, & Belzer, 2010); using drugs and/or sex-related drugs, as well as alcohol, with sexual partners prior to sexual intercourse (Binson et al., 2001; Reisen, Iracheta, Zea, Bianchi, & Poppen, 2010; Reisen, Zea, Bianchi, & Poppen, 2011); and whether communication regarding safer sex took place prior to sexual encounters (Greene & Faulkner, 2005; Gorbach & Holmes, 2003; Noar, Carlyle, & Cole, 2006). Despite this evidence, few have examined partner influence on condom use across differing social environments. Even fewer studies have investigated how condom usage varies by partner characteristics.
Despite more than a decade of HIV-prevention efforts, there is currently no evidence of reduced HIV prevalence among MSM and male-to-female transgender women, hereafter referred to as transgender women, in Indonesia. An analysis of a 2011 Integrated Biological and Behavior Survey, for example, reported that HIV prevalence among MSM rose from 5% in 2007 to 12% in 2011. Meanwhile, HIV prevalence among transgender women remains high, with the trend essentially unchanged from 24% in 2007 to 23% in 2011 (Ministry of Health of the Republic Indonesia [MOH], 2012). Furthermore, HIV prevalence differs across geographic areas in Indonesia. The capital city, Jakarta, has the highest HIV prevalence rate within the country, 17% and 30% for MSM and transgender women, respectively (MOH, 2012). Research focusing on these groups in Indonesia also confirms increasing evidence of HIV-risk behaviors (Morineau et al., 2011; Pisani et al., 2004; Prabawanti et al., 2011).
This article examined condom usage and correlates, with particular interest in differences across several high-risk groups, namely MSM, who congregated and/or sought sexual partners at entertainment places (EPs; e.g., discotheques/dance clubs/karaoke bars), MSM who worked at massage parlors (MPs), and transgender women who congregated and/or sought sexual partners in street/park (S/P) locations, in Jakarta, Indonesia. It has been well documented that venues where MSM and transgender women congregate and/or meet their sexual partners are also associated with HIV infections (Binson & Wood, 2003; Frankis & Flowers, 2009; Grov, 2012; Mimiaga et al., 2011; Phillips et al., 2011; Reidy et al., 2009). These include open public spaces (e.g., “cruising areas,” parks) and commercial settings (e.g., gay bathhouses, saunas, sex clubs). Research has concluded that HIV risks also differ by venue. For example, several studies have concluded that public settings carry a lower risk of HIV transmission compared with commercial settings (Binson et al., 2001; Reisen et al., 2010; Reisen et al., 2011). Certain behaviors, such as the use of drugs, including sex-related drugs as well as alcohol, with sexual partners prior to sexual intercourse have also been identified as factors that may contribute to the likelihood of patron’s HIV-risk behavior in commercial sex settings (Binson et al., 2001; Reisen et al., 2010). Thus, differences in condom use by congregate venues may be observed.
In the face of high HIV-risk behaviors and HIV cases within these population groups, better documentation of HIV risks is needed to prevent further HIV transmission. A multilevel approach that allows simultaneous examination of multiple determinants of MSM and transgender women’s condom use, including variables measured at the individual and at the sexual partner level, was employed in order to enable us to better understand differences in HIV-risk patterns and tailor more effective HIV-prevention efforts.
Method
Design, Sample, and Setting
A cross-sectional design, structured interviews, and hierarchical linear modeling were used to examine condom use among MSM and male-to-female transgender women currently residing in Jakarta, Indonesia. Information about respondent sexual partners was gathered by using an egocentric design (Morris, 2004) in which their sexual partners were neither traced nor enrolled in the study. Contextual information about each type of respondent was obtained through systematic field observations of the sites and informal interviews with respondents and managers/owners.
Ethics approval for the study was obtained from the Institutional Review Board, University of Illinois at Chicago. Local approval was also obtained from the Research Ethics Committee of Atma Jaya Catholic University, Jakarta, Indonesia.
Participants of the study were (a) men who have sex with other men either for commercial gain or as a matter of sexual preference (Morineau et al., 2011; Pisani et al., 2004) or (b) persons self-identified as transgender women (a person who is biologically male and meets the definition of transgender if they cross-dress, adopt the behavior and societal roles of females, and socialize regularly with fellow transgender in definable communities; De Santis, 2009; MOH, 2012; Prabawanti et al., 2011); and who (c) reported residence in Jakarta and surroundings for the previous 6 months; (d) were 18 years and older; (e) congregated and/or sought sexual partners at one of the following venues—discotheques/dance clubs/karaoke bars, massage parlors, streets and/or parks in Jakarta; and (f) were willing to provide oral informed consent to participate in the study.
MSM in Indonesia consist of both gay-identified individuals and nongay-identified MSM who may also report sex with women (Morineau et al., 2011; Pisani et al., 2004). Most transgenders are male-to-female. They are known as waria in Indonesia. This term is a combination of the Indonesian words wanita (female) and pria (male). Several literatures on waria in Indonesia have described characteristics of waria, which include being away from their original home since a young age, having dropped out of school, and having migrated to larger Indonesian cities due to lack of acceptance of their behavior by family and peers (Crisovan, 2006; Prabawanti et al., 2011; Praptoraharjo, 2011).
The characteristics of MSM and transgender women differ by venue where they congregated and/or sought their sexual partners. Characteristics pertaining to each level are described below.
MSM–Entertainment Places (EP)
MSM were recruited from entertainment places (e.g., discotheques/dance clubs/karaoke bars) used by MSM seeking leisure-time activities and sexual partners. A cover charge was required prior to entering these establishments, but rules and prices varied by type of venue. For example, a typical open air traditional dance club would charge a 6,000 rupiah (<$1) entrance fee. Once admitted, patrons could request a song, dance, and/or sing themselves. Likewise, two other dance clubs charged around $8 to $10 for entry. The type of relationships among patrons could be either casual or commercial. Sexual intercourse took place either outside of the venue or at their partner’s rented house or room. By contrast, discotheques bars in Jakarta tended to charge a more expensive entrance fee (between $10 and $15), particularly on weekends and for special events. Generally, while patrons met at the discotheques/bars, the sexual acts occurred elsewhere. Some discotheques, however, provided rooms for sexual acts. Despite being illegal, club drugs (e.g., ecstasy) were commonly consumed by patrons at most discotheques. Finally, karaoke bars usually had a typical traditional karaoke format, consisting of one big hall with a large screen and several chairs, where patrons could sing songs. There were no specific regulations or rules as to who could attend these entertainment venues. As long as they did not commit any criminal activity, all were treated equally. These entertainment venues were typically open from 8 p.m. to around 2 a.m. to 3 a.m.
MSM–Massage Parlors (MP)
MSM who worked at MPs were called “therapists.” MPs were scattered throughout Jakarta and usually differed by physical appearance, number of MSM working at the venue, the prices they charged, and the services they offered. Depending on the service provided (e.g., massage, spa, masturbation, oral, or anal sex), the therapist may receive more of a tip. Most MPs consisted of several rooms where massage services and sexual intercourse took place. Some MPs provided housing for MSM. Some had regular therapists but also permitted freelancers. Local civil society organizations (CSOs) had close collaborations with all participating MPs in this study, thus condoms and lubricants were available and given to therapists along with lotion every time they provided sexual services. Yet the use of condoms depended largely on individual negotiation with each client. Employees also participated in regular HIV/AIDS and sexually transmitted infection (STI) tests conducted by local community health centers and CSOs. An HIV/AIDS regulation regarding condom use signed by the local authorities was posted on the entrance wall at most MPs. Some therapists had a wife and children at home. Usually, their family was unaware about their work. For those who were married, the reason to enter into sex work was typically economic need and lack of other opportunities. The prices charged ranged between 60,000 and 500,000 rupiah ($9-$60). Those who charged for more expensive services usually had similar characteristics as seen in the “bathhouse” in developed countries (Binson et al., 2001; Binson & Wood, 2003; Reidy et al., 2009). Therapists learned about vacant positions at particular MPs through reading ads in local newspapers or on the Internet. Some were also referred by therapists who had worked there. They were subject to fixed working hours and other in-house regulations. Some MPs implemented a contract system with a renewal possibility while others allowed men to work indefinitely as long as they complied with regular working hours and adhered to other rules implemented by the managers-owners.
Transgender–Streets/Parks (S/P)
In addition to having day jobs, such as working as beauticians in nearby salons and/or as street singers, most transgender women, or waria, congregated in the street and dressed up as women to look for clients. Depending on the locations, waria who congregated in streets/parks throughout Jakarta differed also by the prices they charged, type of clients served, and their physical appearance. For example, some charged their clients around 100,000 to 200,000 rupiah ($11-$22). In this case, most of their clients tended to look wealthy. They usually came in private cars where sexual intercourse usually took place. Other waria in different locations were paid around 50,000 to 75,000 rupiah ($6-$11) per sexual act, which took place on a street corner or other dark spaces. Finally, some waria charged their sexual clients around 20,000 to 30,000 rupiah ($3-$4) and provided their services in more primitive environments. In these areas, a surface made of soil (without roof and bed) was available to be rented once clients were obtained. Seeing others engaging in sexual acts was a common phenomenon as each room was only separated by a thin curtain and no roof. The room cost around 5,000 rupiah ($0.5-$1). Once rented, they received a newspaper to cover the floor and a condom that was distributed by local CSOs. If no rooms were available, then the sexual acts occurred standing up. As clean water was not always available during sexual acts in streets/parks, they preferred using hand body lotion as a lubricant because of ease of cleaning by small towel rather than water.
Not all waria necessarily had sex for money; rather they had sex for their own pleasure. Once attracted to a particular client, she may decide to have sex with him for free. Additionally, they had regular partners. Often these partners had a wife and children. Despite being aware that they were being used by men for money, they viewed it as fulfillment of their manhood side to take care of someone. Meanwhile, when they sought clients at night, it was considered fulfillment of their feminine side. Most waria had a good network to help them find a day job and/or to deal with any personal and HIV/AIDS-related issues. There were also key people responsible to maintain and secure the venue. There were no specific regulations or rules as to who visited these venues. As long as they got permission from these key people and/or the owners of the venues and did not commit any criminal activity, they were free to congregate or seek male clients. Their peers usually referred them to new locations. Similar to other high-risk populations (e.g., female sex workers), waria also had become a target of the police or local religious groups who disapproved of prostitution. However, to protect the venue, in some areas, waria were required to contribute additional money to pay the police or religious groups.
There were several kiosks surrounding the locations that sold both traditional and generic liquors along with snacks and cigarettes. Waria usually congregated as a group at kiosks to consume alcohol. In this way, they could be less shy in approaching male clients. Most waria were contacted for HIV/AIDS-related services by local CSOs in their day jobs (e.g., beauty salon) or in their rented homes.
Data Collection
During March to April, 2012, a total of 302 MSM and transgender women were recruited from the following venues: discotheques/karaoke bars/café/pub/dance club (n = 91), massage parlors (n = 97), and streets/parks (n = 114) in five different neighborhoods in Jakarta. The venues and number of participants sampled within each were determined based on information provided by CSOs that currently conduct HIV/AIDS-prevention programs with these groups. Gaining access to target participants was relatively easy. All owners, managers, or persons-in-charge at the venues approached granted permission to conduct the interviews; thus a 100% cooperation rate was attained. Likewise, all participants who were identified and approached at each location during the study period also agreed to participate.
Measures
The dependent variable for this analysis was a dichotomous measure of whether or not respondents used a condom during anal sex with each partner during the past 7 days. The measure was derived from two questions: “Could you list the last three recent people that you had anal sex with in the past 7 days?” and “Did you use a condom when you had anal intercourse with him/her?” (Yes/no). Independent variables were measured at two levels: respondent and sexual partner.
Respondent-level variables included demographic characteristics, coded as (a) respondent type (MSM-EP vs. MSM-MP vs. transgender-S/P); (b) age of respondent (≤25 years vs. >25 years). Although age is conceptually a continuous variable, it is commonly treated as categorical in statistical analyses for substantive and practical reasons. In our case, we truncated the age variables at the respondent-, and later at the sexual partner-level because of their somewhat skewed distributions. We used mean values as approximate cut points; (c) education level (less than or equal to junior high school vs. senior high school and higher); and (d) marital status (never married vs. ever married). History of assault and abuse were measured with the following questions: (a) childhood sexual abuse (CSA), “Do you feel you were sexually abused when you were growing up (e.g., before 18 years)?” (yes vs. no); (b) adult sexual abuse (ASA), “Since you were 18 years old, was there a time when someone forced you to have sexual activity that you really did not want? This might have been intercourse or other forms of sexual activity, and might have happened with lovers or friends, as well as with some distant persons or stranger” (yes vs. no); (c) childhood physical abuse (CPA), “Do you feel that you were physically abused by your parents or other family members when you were growing up?” (yes vs. no); and (d) adult physical abuse (APA), “Since age 18, has anyone ever physically assaulted you or attacked you with a weapon?” (yes vs. no). Having experienced depressive symptoms in the past week was assessed using an established subset of 9 items from the revised Center for Epidemiological Studies–Depression scale (CES-D; Santor & Coyne, 1997; α = .780). The 9 CES-D items were a short version of the 20-item CES-D scale that has been demonstrated to reliably represent the full 20-item scale (Santor & Coyne, 1997). It was chosen to reduce instrument length and to render it more user-friendly. The total possible score range for the 9-item scale was 0-27. Having experienced stigma and discrimination related to sexual orientation was measured by using a subset of questions pertaining to enacted stigma and discrimination that were adapted from a study of stigma and discrimination among MSM in the United States and China (Diaz et al., 2001; Neilands et al., 2008; α = 0.619). Higher summated scores represented greater levels of depression and experiences of stigma and discrimination, respectively. Knowledge about HIV transmission was measured using a summated index based on the total number of correct responses to a series of 11 questions adopted from previous study of MSM and transgender women in Indonesia (MOH, 2012; α = 0.893). The 11 questions addressed a combination of facts and myths regarding how HIV was transmitted. Examples included statements, such as “sexual intercourse without a condom” and “drinking out of someone else’s glass” could transmit HIV or not, and so on. The precise wording of all items is available from the authors.
The sexual partner–level variables were derived by asking respondents to nominate their last three sexual partners whom they had anal sex with during the past 7 days. Respondents were asked to provide information on the attributes of each partner, partner-specific sexual behaviors, and whether they used club drugs, drank alcohol, and/or used sex drugs with each partner. The variables were coded as: (a) type of partner (regular, casual, or commercial); (b) partner age (≤35 years vs. >35 years); (c) use of club drugs, such as methamphetamine and/or ecstasy before sex (yes or no); (d) having drank alcohol before sex (yes or no); and (e) use of sex drugs, such as Viagra, before sex (yes or no).
Data Analysis
A two-level hierarchical analysis with a random intercept was employed to account for dependencies in the partners’ behavior (use of condom) nested within individual MSM and/or transgender female respondents. The hierarchical linear and nonlinear modeling (HLM) 6.0 software by Raudenbush, Bryk, Cheong, Congdon, & Du Toit (2002) was used. The model controls for variables that were selected as potential predictors of condom use based on previous research and literature review of factors associated with condom use within these groups.
Results
Of the 302 respondents who reported having anal sex in the past 7 days, 131 reported at least three sexual partners during that time, thus contributing 393 observations to the partner data set; 77 respondents reported having two sexual partners in the past 7 days, thus contributing 154 observations. Finally, 94 respondents reported having only one sexual partner during that time, which contributed 94 observations to the partner data set. In total, 641 observations were thus available for analysis in the sexual partner data set. These sexual partners were linked to the 302 respondents.
Descriptive Results of Respondent Characteristics
Of the 302 respondents interviewed, the percentages of MSM-EP, MSM-MP, and transgender-S/P were 30%, 32%, and 38%, respectively (see Table 1). More than half of the respondents were older than 25 years (61%), and 65% had attained more than a senior high school education. Meanwhile, more than three quarters reported never having been married (84%). Less than half reported having experienced child and adult sexual abuse (46% and 41%, respectively). Less than a quarter reported having ever experienced child and adult physical abuse (24% and 20%, respectively). On average, respondents were able to correctly answer more than half of the 11 knowledge questions concerning HIV transmission (M = 8.3, SD = 2.4). Mean CES-D and stigma and discrimination scores were 9.0 (SD = 5.8, range = 0-27) and 8.7 (SD = 2.7, range = 6-21), respectively.
Descriptive Statistics of Sexual Partner– and Respondent-Level Variables.
Note. MSM-EP = men who have sex with men recruited from entertainment places; MSM-MP = MSM recruited from massage parlors; TG-SP = transgender women who congregated and/or sought sexual partners on streets/parks.
Sexual Partners Characteristics
In terms of respondent’s sexual partner characteristics, the overall rate of condom use was 79%. That is, condoms were reported to have been used during 79% of the sexual encounters involving anal sex that were reported to have taken place by respondents during the past 7 days. The rate of condom use varied significantly across MSM-EP, MSM-MP, and transgender-S/P groups (66%, 84%, and 83%, respectively), χ2(df = 2) = 21.59, p = .000. Meanwhile, the rates of alcohol use, club drug use and the use of sex drugs (e.g., Viagra) were 15%, 9%, and 16%, respectively for the sample as a whole. Finally, of the 641 total sexual encounters reported, more than half of the sexual partners were commercial (65%) and younger than 35 years (74%; see Table 1).
Predictors of Condom Use at Respondent and Sexual Partner Level
Results of a random intercept two-level hierarchical model indicated some significant predictors of condom use (Table 2). At the respondent level, type of respondent was not associated with condom use. However, several respondent demographic and psychological characteristics were associated with condom use. For example, a greater likelihood of condom use was reported among those who had attained higher education (odds ratio [OR] = 1.88, 95% confidence interval [CI] = 1.10-3.24). Lower condom use was associated with reporting more depressive symptoms in the past week (OR = 0.94, 95% CI = 0.91-0.98), and having a history of physical abuse during childhood (OR = 0.34, 95% CI = 0.19-0.61). Finally, knowledge of HIV transmission pathways had a marginally significant effect on increasing condom use, and adult physical abuse had a marginally significant effect on decreasing condom use (OR = 1.09, 95% CI = 0.99-1.20, p = .056; and OR = 0.56, 95% CI = 0.31-1.00, p = .05, respectively).
Multilevel Logistic Regression Results for Condom Use Among Sexual Partners of MSM and Transgender Women.
Note. MSM, men who have sex with men; MSM-EP = MSM recruited from entertainment places; MSM-MP = MSM recruited from massage parlors; TG-SP = transgender women who congregated and/or sought sexual partners on streets/parks.
At the sexual partner level, the likelihood of condom use was higher with casual (OR = 2.30, 95% CI = 1.19-4.41) and commercial (OR = 5.68, 95% CI = 3.15-10.23) partners, compared with regular partners. Those who reported using club drugs with their sexual partners before sex were less likely to report using condoms (OR = 0.25, 95% CI = 0.12-0.53).
Discussion
Our study indicates that reported condom use was high among MSM and transgender women in Jakarta. The fact that HIV prevalence is also high among these groups, despite exposure to multiple HIV programs, suggests the possibility that respondents may be overreporting condom use (MOH, 2012; Prabawanti et al., 2011). It is known that some survey methodologies can minimize the social desirability pressures placed on respondents. For example, research in other settings has documented fewer socially desirable responses to self-administered questionnaires (Tourangeau, Risp, & Kenneth, 2000). However, most of the survey research conducted in Indonesia has employed interviewer-assisted interviews. It is also known that what is perceived as socially desirable is culturally conditioned (T. P. Johnson & Van De Vijver, 2003). Additional research to document the degree to which HIV research within the Indonesian context may be vulnerable to social desirability bias should be a future priority.
Multilevel analyses identified associations between several psychological characteristics and condom use among these groups that were consistent with other studies. For example, depressive symptoms in the past week were associated with reported condom use, suggesting that negative affect might have contributed to poor decision making regarding personal safety (Clements-Nolle et al., 2008; Reisner, Mimiaga, Skeer, & Mayer, 2009; Salomon et al., 2009), thus limiting their ability to practice safer sex. Likewise, histories of physical abuse during childhood and adulthood (Kosenko, 2011; Lombardi et al., 2001; Maman et al., 2000) were observed in these analyses. These variables are also known to affect one’s self-esteem and ability to effectively negotiate safer sexual practices, thus also engendering risk behavior. Knowledge of HIV transmission pathways was also marginally associated with greater condom use, consistent with other research in Indonesia (MOH, 2012; Prabawanti et al., 2011). Accurate HIV/AIDS transmission and prevention knowledge can also be expected to empower individuals and increase the likelihood they will take action to modify their behavior.
Furthermore, our findings revealed that condom use was more common among casual and commercial sex partners compared with regular partners. This is consistent with prior research that has concluded that MSM and transgender women are less likely to engage in HIV-risk behavior when their partners are commercial and/or casual (De Santis, 2012; Mi, Wu, Zhang, & Zhang, 2007; Morineau et al., 2011; Nemoto et al., 2004; Operario et al., 2011; Praptoraharjo, 2011; Sethi et al., 2006). This might be due to the far stronger familiarity, trust, and relational bonds that are typically established between regular partners, which might render the perceived need for protection less important or necessary. Further examination of the full range of sexual relationships by partner types and characteristics, particularly those that have been reported to influence condom use, should be considered (Clements-Nolle et al., 2008; De Santis, 2012; Eaton et al., 2009; Gorbach & Holmes, 2003; Mi et al., 2007; Nemoto et al., 2004; Praptoraharjo, 2011).
Our study also demonstrates that the use of club drugs with sexual partners is associated with lower likelihood of condom use, another finding consistent with prior studies (Boeri et al., 2004; Buchbinder et al., 2005; Diaz et al., 2005; Lyons et al., 2010). The use of club drugs (e.g., methamphetamine) is worrisome because of the fact that they are commonly used in conjunction with sexual risk taking and thus may be a proxy indicator of high-risk sexual behaviors (Carey et al., 2009).
Our systematic field observations revealed a number of characteristics that further place our sample at increased risk of acquiring and transmitting HIV. These characteristics differed by type of respondent. For example, the availability of alcohol and/or drugs in entertainment places could place MSM-EP at increased risk. It should also be noted that MSM-MP who exchanged sex for money were not necessarily “gay” in all cases. Some were married and had children, a consistent finding with other MSM studies in Indonesia (Morineau et al., 2011; Pisani et al., 2004) and in other Asian countries (Asthana & Oostvogels, 2001; Choi et al., 2003; Colby, 2003). This was of additional concern given the risk of transmission to regular partners. Finally, as also seen in other studies among transgender women in Indonesia (Crisovan, 2006; Prabawanti et al., 2011; Praptoraharjo, 2011) and elsewhere (Clements-Nolle et al., 2008; De Santis, 2009; Nemoto et al., 2004), most transgender women in our study had a day job, as well as routine exposure to stigma, discrimination, and violence. The fact that they had double roles as men and women could also diversify the characteristics of their sexual partners, thus increasing their likelihood of both acquiring and transmitting HIV. Furthermore, the fact that there was poor hygiene, no proper place to have sex, and no clean water in some transgender locations could add to their risk of contracting STIs. In addition, transgender women preferred using hand/body lotion as a lubricant because of ease of cleaning by small towel rather than water. This is also worrisome, as they may contract other STIs and/or other genital infections when using nonregular lubricants. Indeed, previous research has also reported high prevalence of rectal gonorrhea and/or chlamydia among transgender women in Indonesia (MOH, 2012; Pisani et al., 2004; Prabawanti et al., 2011).
Given the key role of managers/owners at places patronized by MSM, their support is important in encouraging HIV risk reduction (Woods, Euren, Pollack, & Binson, 2010). Condom use promotion and marketing should be emphasized within all venues regardless of their physical characteristics. Easy access to condoms through restroom dispensers and other means might also further encourage their use. It would seem worthwhile to profile opportunities for encouraging HIV-risk reduction as being “good for business” when seeking support from managers/owners of these venues.
Our findings also suggest emphasizing messages that HIV risk reduction measures should be undertaken regardless of the types of relationships individuals have with their sexual partners. Promoting and encouraging dyadic communication, particularly on HIV-related issues and condom use prior to sexual intercourse, should be considered. Indeed, several studies among both heterosexual and homosexual populations have confirmed that the greater the communication regarding safer sex within sexual dyads, the higher the level of condom use observed (Greene & Faulkner, 2005; Gorbach & Holmes, 2003; Noar et al., 2006).
Finally, HIV-prevention efforts should also be extended to include counseling for depression and history of physical abuse. Collaboration efforts with local community health centers for HIV and STI screening are needed. Proper screening for depression and background history to identify and address physical and sexual abuse experiences can be done prior to HIV testing.
Several limitations should be acknowledged. Most important, as already discussed, sexual behavior was self-reported and may be subject to errors of memory and/or willingness to report accurately. Furthermore, interviewers in this study were staff members of CSOs who provided HIV/AIDS-related services to this community. This may have influenced respondent answers. To ensure lack of bias, the strategy employed sought to employ interviewers with no preexisting relationship or contact with the interview venue in question. In some cases, though, respondents were nonetheless aware that the person interviewing them was employed by a CSO. Interviewer status as a CSO employee may have influenced several HIV/AIDS-related questions, such as claiming adoption of the correct HIV risk reduction measures. Sampling bias may also have occurred as a result of the venue sampling strategy employed. The final sample may not be representative of MSM and transgender populations in Jakarta beyond those who frequent the specific venue types where field work was done. Only two types of MSM were included in the sample; those who congregated and/or sought sexual partners at entertainment places (e.g., discotheques/dance clubs/karaoke bars) and those who worked at massage parlors. Partner data were based entirely on proxy reporting (by the participants) and are also subject to reporting errors, particularly on questions related to sexual behavior and condom use. Finally, although this study is efficient at identifying associations, it cannot establish cause and effect due to its cross-sectional design.
The study also has several important strengths. These findings contribute to the existing literature on factors associated with condom use among MSM and transgender women in a high-risk Third World setting. In addition, the utilization of multilevel methods has yielded more robust findings regarding condom use variability by type of MSM and transgender women. These findings suggest the potential importance of designing specific target interventions for addressing the HIV epidemic within these vulnerable and stigmatized groups. Further research that includes more detailed examination of the effect of specific venue types and characteristics should be conducted to address specific venue-based interventions. Continuing to improve the available methodologies for measuring sexual network patterns will also be an important next step.
Footnotes
Acknowledgements
We would like to thank Yayasan Srikandi Sejati, Yayasan Intermedika, and LPA Karya Bakti, who assisted the senior author with data collection.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by the International AIDS Society, the National Institute on Drug Abuse, and the Eunice Shriver National Institute of Child Health and Development.