Access to community-based HIV services among transgender women in Cambodia: findings from a national survey

Background Globally, the prevalence of HIV among transgender women is more than 40 times higher than the prevalence in the general reproductive-age adults. They also face intersecting barriers to health, social, and legal services due to their hidden and stigmatized nature. Despite the particular needs, data regarding the access to services among transgender populations is scant globally. This study aims to identify characteristics of transgender women in Cambodia that may determine their accessibility to community-based HIV services. Methods In the National Biological and Behavioral Survey 2016, a structured questionnaire was used for face-to-face interviews with 1375 sexually active transgender women recruited from the capital city and 12 other provinces using the Respondent-Driven Sampling method. Weighted multivariate regression analysis was conducted to explore factors associated with access to community-based HIV services. Results The mean age of the participants was 25.8 years (SD = 7.1), and 45.0% reported having received at least one community-based HIV service in the past three months. Compared to participants who reported not having been reached by any community-based HIV programs, participants who reported having been reached by the programs were significantly more likely to reside in an urban setting (AOR = 1.41, 95% CI = 1.01–1.96), to have used gender-affirming hormones (AOR = 1.50, 95% CI = 1.17–1.92), to have been tested for HIV in the past six months (AOR = 7.42, 95% CI = 5.78–9.53), and to have been arrested by police or other authorities because of their transgender identity (AOR = 1.55, 95% CI = 1.03–2.33). Participants who reported having been reached by community-based HIV programs were significantly less likely to report being in a receptive role (AOR = 0.34, 95% CI = 0.15–0.82), to use condoms consistently with non-commercial male partners (AOR = 0.72, 95% CI = 0.55–0.94), and to perceive that their co-workers were not supportive regarding their transgender identity (AOR = 0.57, 95% CI = 0.44–0.98). Conclusions Despite the extensive expansion of community-based HIV programs, less than half of transgender women in this national survey had access to the services. Innovative strategies and culturally sensitive interventions should be put in place to reach and respond to the needs of sub-groups of transgender women who are less likely to be reached by the existing traditional approaches.


Background
According to the World Health Organization (WHO), "transgender people" is an umbrella term for all people whose internal sense of their gender (gender identity) is different from the sex they were assigned at birth [1]. Transgender people choose different terms to describe themselves. There are some transgender people who do not identify themselves as either male or female, but rather outside of a gender binary. In some cultures, specific indigenous terms are used more commonly to describe transgender people or those who identify as a third sex [1]. Transgender women, who are biologically male at birth and self-identify as female [1], are considered at a heightened risk of acquiring HIV and other sexually transmitted infections (STIs) but are strikingly underrepresented in HIV programs and national surveillance systems.
Globally, the prevalence of HIV among transgender women is estimated to be 19.1% (95% CI: 17.4-20.7), which is more than 40 times higher than the prevalence in the general reproductive-age adults [2]. Transgender women experience multiple, intersecting socio-structural risks that increase their vulnerability to HIV, including stigma and discrimination, gender-based violence, lack of social and legal recognition of their affirmed gender, and exclusion from employment and educational opportunities [3,4]. These situations may expose them to subsequent mental health problems or vice versa [5][6][7]. Among transgender women living with HIV, these systems of oppression can seriously compromise their sustained access to life-saving treatment and care [3,8].
The WHO recommends community-based HIV and legal services as a way to reach and respond to the intersecting needs of key populations, including transgender women [9]. This human rights-based approach posits that programs must recognize and respond to the multi-faceted needs arising from transgender women's social position [4,8,10]. It also recognizes that HIV services need to be located within transgender women's communities or in other sites where they feel comfortable [9,11,12]. Furthermore, there has been an increased demand for integrated services that address not only HIV but also other health issues as well as legal and psychosocial needs [12,13].
In addition to human rights-based arguments, a strong meta-analytic evidence base suggests that community-based HIV services hold promise for improving HIV prevention, treatment, and care. A global systematic review of 117 studies, including 864,651 participants, found that relative to facility-based approaches, community-based programs increased uptake of HIV testing and counselling (HCT), first-time testing, and the proportion of testers with CD4 counts above 350 cells/μl [14]. Another systematic review of 22 cohort studies and randomized control trials, including 97,657 participants, found that community-based HIV programs resulted in comparable ART adherence, virological suppression, and all-cause mortality outcomes as facility-based services [15]. Moreover, community-based programs had the added benefit of higher rates of treatment engagement and cost-saving [4,15].
However, little is known about the extent to which transgender women engage with community-based programs [16]. Part of the reasons may be a dearth of representative, national-level data on transgender women. Existing data have been collected using inconsistent definitions of transgender, frequently classifying transgender women as a subcategory of men who have sex with men [17]. This approach erases transgender identities, disregards their specific needs and undermines epidemiological data. Moreover, to our knowledge, no studies have examined the health and sociodemographic characteristics of transgender women who access and those who do not access community-based HIV, other health, and legal services.
This paper aims to address these notable gaps using data from a large-scale national bio-behavioral survey of transgender women in Cambodia. This study has two objectives: (1) to examine the extent to which transgender women in Cambodia access community-based HIV and legal services and (2) to identify the characteristics of transgender women who do and do not access community-based HIV services so as to inform targeted services for this underserved population.

Study sites and participants
The National Integrated Biological and Behavioral Survey among transgender women (TG-IBBS 2016) was conducted in the capital city of Phnom Penh and 12 other provinces with high burden of HIV and a large population of transgender women. These provinces were purposively selected by the research team and supported by the national HIV technical working group. More than 70% of the Cambodian population resides in the 13 provinces out of 24 provinces and the capital in Cambodia. According to the national program data from the National Center for HIV/AIDS, Dermatology and STD (NCHADS), in 2015, the selected sites consisted of 21 operational districts (ODs) with the highest-burden of HIV, and more than 80% of men who have sex with men and transgender women in Cambodia.
The Respondent-Driven Sampling (RDS) method was used to recruit participants from 20 specific locations (six locations in Phnom Penh and 14 locations in the remaining 12 provinces). The number of locations was determined based on the estimated population of transgender women and the proportion of the required sample size in each study site. Individuals would be included in the survey if they: (1) were biologically male at birth and self-identified as female, (2) were aged 18 years or older, (3) reported having sex with at least one man in the past 12 months, (4) were able and willing to provide written consent to participate in the study, and (5) were able to communicate in Khmer, the national language of Cambodia. Full details of the methodology have been reported elsewhere [18].

Training and data collection
A three-day training was provided for all research team members. The training included a review of the study protocol, interview techniques, and confidentiality and privacy protection. It also entailed interview practice through role-play and study procedures through a pilot study. Throughout the data collection, daily review sessions were conducted with interviewers to review progress and communicate any problems that may emerge during the interviews.
Two data collection teams, with eight members each, were formed. Each team included one field supervisor, five interviewers, one counselor, and one lab technician from the Provincial AIDS and STI Program. A peer outreach worker working for transgender women in the area also joined the team as a facilitator. The field supervisor performed participant eligibility screening. A unique personal identification number was assigned to each consenting participant. The number was used to link all data collected from each participant, but it was not linked with any personal identifiers of the participants. Written informed consent was obtained after the counselor informed the participant of the study details. After obtaining the consent, the participant was interviewed in a private room. Each participant received US$4 in cash to compensate for their transport and time and a package of three condoms. In addition, seeds and recruiters received $2 for a successful referral.

Variables and measurements
A structured questionnaire was initially developed in English and then translated into Khmer. The Khmer questionnaire was then back-translated by another translator to ensure that the "content and spirit" of every original item was maintained. It was validated through consultative meetings with key stakeholders and representatives of transgender women, pretested with 20 transgender women in Phnom Penh, who were later excluded from the study, and modified based upon comments and suggestions from the pretest and meetings.
Most of the socio-economic and health risk behavior items were adapted from the 2014 Cambodia Demographic and Health Survey [19] and our previous community-based surveys among HIV key populations [20][21][22]. Socio-demographic characteristics included age, study sites (urban, rural), gender identity, marital status, main occupation, average monthly income in the past six months, completed years of formal education, duration living in the current city, and perceived family economic status. Information regarding gender expression and utilization of gender-affirming hormones and surgeries was also collected. Regarding HIV risks, information collected included sexual behaviors in the past three months such as number of sexual partners, condom use with different types of partners, experiences of STI symptoms, and care seeking behaviors. For transgender women who reported living with HIV, self-reported data on HIV care and treatment were collected.
Regarding access to community-based HIV programs, participants were asked whether they had received the following services in the past three months from a community-based organization: (1) HIV and sexual reproductive health education, (2) condom and lubricant distribution, (3) screening for HIV and STIs, (4) legal support services, (5) other health services such as HIV and STI care and treatment, referral for other health services, etc., and (6) online services developed for transgender women (websites, Facebook pages, hotline calls, etc.). A dichotomous variable was created to capture access to community-based HIV services defined as having been reached by a community-based HIV program through any one of the services in the past three months versus none.
Exposure to gender-based discrimination and violence was measured using items adapted from previous studies [23,24]. Participants were asked about different forms of discrimination and violence they had experienced. These included whether they felt that their co-workers or classmates were supportive regarding their transgender identity and had experienced problems such as difficulties in getting a job, losing a job, having been denied or thrown out of a housing, and difficulties in getting HIV or other health services and thought it was because of their transgender identity or expression. Participants were asked about their experiences in different forms of violence such as having been physically or sexually abused, been arrested by authorities, dropped out of school, and been fearful of being arrested by police or other authorities because of their transgender identity or expression.

Data analyses
Double data entry was performed using EpiData version 3 (Odense, Denmark). STATA Version 12.0 (Stata Corp, Texas, United States) was used to conduct the statistical analyses. All analyses were estimated with sampling weights that corrected for nonresponse and sample design [25]. Bivariate analyses were run, using χ 2 test (or Fisher's exact test when a sample size was smaller than five in one cell) for categorical variables and Student's t-test for continuous variables to compare socio-demographic characteristics, HIV risk behaviors, and experiences of gender-based discrimination and violence among transgender women who reported having and having not been reached by community-based HIV programs. To explore independent factors associated with access to community-based HIV programs, a weighted multivariate logistic regression model was constructed, controlling for potential confounders. All variables significantly associated with access to community-based HIV programs in the bivariate analyses at a level of p-value < 0.05 were simultaneously included in the model. Age, education level, and income were included regardless of the significance level because of their epidemiological importance. Adjusted odds ratio (AOR) were obtained and presented with 95% confidence interval (CI) and p-values.

Ethical considerations
The National Ethics Committee for Health Research (NECHR) of the Ministry of Health, Cambodia (No. 420 NECHR) approved the study protocol and materials. A written informed consent was obtained from each participant, and participants were informed that they could decline or withdraw from the study at any time. Privacy and confidentiality were protected by having the interviews conducted in a private room and removing all personal identifiers from all study documents.

Community-based HIV programs
The survey sample included 1375 transgender women with a mean age of 25.8 years (SD = 7.1). Table 1 shows the proportion of transgender women living in urban and rural areas who had access to different communitybased HIV services in the past three months. Of the total, 45.0% of the participants reported having received at least one community-based HIV service in the past three months. The most common services they had received included condom and lubricant distribution (38.8%), HIV and sexual and reproductive health education (35.6%), online HIV-related information (35.1%), and screening for HIV and STIs (26.9%). Table 2 shows that the majority of the participants lived in an urban area (83.3%) and were not married or living with a sexual partner (78.1%), and 53.0% were in the age group of 18-24 years. Their median average monthly income in the past six months was US$150 (interquartile range = 90-200), and the mean completed years of formal education they had attained was 9.0 (SD = 3.4). Around one-third (35.1%) were a hairdresser or beautician. Less than half (45.0%) reported having used gender-affirming hormones, and 9.3% having had gender-affirming surgery. Significant differences between participants who reported having and having not been reached by community-based HIV programs were found in type of communities (rural or urban), length of stay in the current location, monthly income, main occupation, frequency of expression as female, number of other transgender women they knew, and use of gender-affirming hormones and surgery.

HIV risk behaviors and HIV testing
According to Table 3, 8.6% of the participants reported having sex with women, and 1.6% reported did so in exchange for money or gifts in the past three months. The majority of respondents (87.5%) reported being in a receptive role in anal intercourse with men, and 81.6% having sex with men not in exchange for money or gifts in the past three months. More than one-third (37.8%) reported having sex with men in exchange for money or gifts, with 60.0% reporting always using condoms with this type of partners. Regarding HIV and STI services, 14.0% had been diagnosed with an STI, and 60.1% tested for HIV in the past six months. Significant differences between the participants who reported having and having not been  reached by community-based HIV programs was found in sexual intercourse with women, usual roles in anal sex with men, condom use, and HIV testing.

Gender-based discrimination and violence
It was common for transgender women in this study to experience different forms of discrimination and violence because of their transgender identity (Table 4). Common experiences included difficulties in getting a job (42.0%), sexual abuse or assault (39.1%), feeling fearful of being arrested by police or other authorities (24.9%), having lost a job (24.3%), physical abuse (23.6%), having been denied or thrown out of a housing (18.8%), and feeling that their co-workers or classmates were not supportive regarding their transgender identity (10.1%). Significant difference between participants who reported having been and having not been reached by community-based HIV programs was found in the levels of support from their co-workers or classmates regarding their transgender identity, difficulties in accessing HIV services, physical and sexual abuse and assault, and getting arrested by police or other authorities.

Factors associated with access to community-based HIV services
As shown in Table 5

Discussion
One objective of this study is to examine the extent to which transgender women access community-based HIV and legal services in Cambodia. Globally, the needs to offer and improve differentiated services for key populations most affected by HIV, including transgender women, are growing [9]. Findings from this study indicate that almost half of transgender women respondents had been reached by various community-based HIV services to a varying degree. In line with findings from other studies on transgender women and other key populations [14,15,26] and the recommendations from the WHO [9], the findings suggest that community-based services are an important gateway to healthcare services and service uptake for transgender women. Besides, access to community-based, gender-affirming, and trans-specific services are known to increase HIV prevention service uptake and to enhance their confidence and involvement in HIV prevention in other studies [23,[27][28][29][30]. Compared with other facility-based approaches, community-based programs have been shown to improve HIV prevention and treatment and to be more cost effective [14,15]. This point was beyond the scope of this research; however, findings from our recent study indicated that transgender women were more likely to use NGO-run and community-based health facilities for HIV testing but to use other health facilities for STI treatment [18]. The effectiveness and efficiency of a health facility, including community-based programs, should thus warrant further investigation. It is worth-noting that more than half of the respondents in this national survey had not been reached by any community-based HIV programs in the past three months. Besides, among those who accessed the community-based services, the use of legal and other health services (besides HIV and STI) and new media platforms designed specifically to service their needs had been very low. Given the limited access as well as (cost) effectiveness and efficiency of community-based services, more efforts are needed to better understand the needs for community-based intervention programs among transgender women, including exploring reasons why the women are not reached, why they use a facility or not, which services they consume or not, and which approaches are more applicable and yield better impacts. Such information would be useful for designing the interventions to attain optimal results. This study also offers unprecedented data on the characteristics of transgender women who had and who had not been reached by community-based HIV programs, which is the main and second objective of our study. The findings from the multivariate logistic regression analysis suggest that transgender women who had been reached by community-based HIV programs were more likely to live in urban areas, to use gender-affirming hormones and surgeries, and to have been arrested by police or other authorities due to their transgender identity. They were also more likely to have supportive friends and co-workers of their Values are numbers of subjects (%) for categorical variables *Defined by having been reached by a community-based HIV program with any one of the services † Chi-square test (or Fisher's exact test when a cell count was smaller than 5) was used gender identity and expression. They were less likely to play a receptive role in sexual intercourse and to consistently use condoms with non-commercial male partners [20]. Why and how they possess these characteristics and whether they are associated with or caused by the use of community-based health facilities were not explored in our study and are worth in-depth examination to inform policy formulation and project/program intervention. The possession of these characteristics by the transgender women is not unique to Cambodia, and many of these traits and practices can be detrimental to their health and well-being. Similar to findings from Thailand, which reported that Thai transgender women had limited access to surgery (11%) [31], this study found that few Cambodian transgender women experienced gender-affirming surgerywhich may be linked with their low income as discussed below and the high cost of such surgery. However, a higher number (45%) of transgender women used gender-affirming hormones, and transgender women who had been reached by community-based services were more likely to use such hormones. Inappropriate use of injecting hormones could increase vulnerability to HIV and other blood-borne viruses [31][32][33]. Our recent study found that the risk of HIV infection was more than four times higher among transgender women who self-injected hormones compared to that among transgender women who did not [34].
That transgender womenwhether they access community-based services or notare likely to be harassed or even arrested by law enforcement officers, particularly for their gender identity and expression, is rather a global phenomenon and has been documented in other studies elsewhere [4,35]. Together with the fact that they were less likely to receive legal services from community-based organizations and perhaps other facilities too, the alarming situation reinforces the need for comprehensive intervention programs that may include provision of legal services and training and promotion of advocacy skills and capacity [4,36]. Such programs would require involvement of broader stakeholders at different levels.
A significantly higher proportion of transgender women who had been reached by community-based HIV programs perceived that their co-workers or classmates were supportive regarding their transgender identity. Whether this is a causality or a mere correlation between access to community-based programs and peer support was not covered in the study, and this requires further examination. Apart from the feeling of support as illustrated in the regression model, the findings from our descriptive results show that a fairly high proportion of transgender women reported experiencing different forms of gender-based discrimination and violence. Previous studies in other settings have shown that stigma and discrimination and social exclusion are common among transgender women [37][38][39][40]. These may lead to increased unmet needs for healthcare services and hinder their access to prevention programs that in turn extend the vulnerability to HIV infection among this key population. Gender-based discrimination and violence may cause Defined by having been reached by a community-based HIV program with any one of the services in the past 3 months psychological pressure, mental problems or even suicide [37][38][39][40]. They are also known to be associated with inconsistent condom use in anal intercourse and sex work among transgender women [41][42][43].
Our study indicated that transgender women who had been reached by community-based HIV programs were less likely to report always using condoms with noncommercial partners. The finding is inconsistent with results found in a study of transgender women in India, which reported that transgender women who had access to a community-based program were more likely to use condoms with both regular and casual partners [4]. We were not able to account for the difference in this outcome. However, the high rate of inconsistent condom use in different types of relationships among transgender women is a global phenomenon and has been widely documented in other studies, and this was true even among transgender women who had been reached by community-based HIV intervention programs. These practices place them in an increased risk of contracting HIV and other STIs. The inconsistent condom use can partly be explained by the trust in sexual partners, a sense of freedom and intimacy, as well as low self-esteem and power imbalance in condom use negotiation given their gender identity [41,44,45].
Two key findings from the descriptive results of the characteristics of transgender women are worth-mentioning. Despite the slightly different income between those who had and who had not been reached by community-based HIV programs, they were more concentrated in the informal economy, which is associated with little social safety net and low pay. The low socio-economic status, such as destitution and job insecurity, among our study participants is in line with the earning and living conditions of their peers elsewhere [23,27,28,41]. This finding suggests that community-based programs may consider including non-HIV related services such as job referral training and placement programs [46] to improve their livelihood and career prospects. Another important point is that the number of transgender women within the network is quite large. While this may have good implications for social and peer support and service uptake, findings from other contexts indicated that such a wide network can have implications for HIV and STI transmission as well as commitment of sex work as a means of livelihood and a choice [31].
Finally, our study confirms the promising use of snowballing and peer-driven intervention for HIV case detection and its implication in HIV prevention, care, and treatment among transgender women. With the use of the Respondent Driven Sampling method, we were able to reach hundreds of transgender women (more than half of the total sample) who never used the community-based HIV services. Peer-driven interventions are known to reach wider targets and to produce better outcomes than a traditional approach in case detection and preventive service uptake and retention in care and treatment [4,[47][48][49].
Overall, there is a need to better understand transgender women's access to health and non-health services, especially through the community-based programs, given the limited availability of information and research on this key population and the high prevalence of HIV among this group. Besides, given the (cost) effectiveness, efficiency of community-based programs, the rise of new and social media, and the need for community empowerment and engagement, a holistic approach to HIV prevention and primary care in community-based settings is preferable and can yield better intervention results [26,50,51]. Such an integrated approach may consider provision of services beyond HIV and STI prevention, care, and treatment to include other customized services such as job referral, vocational/technical training, financial aid, legal services, psychological support and counseling, and the use of news and social media for health promotion purposes and beyond.
This study has a number of limitations. First, the findings should be interpreted with caution given the cross-sectional nature of the data. The order of effects remains unknown, and it was not meant to determine causality. Further, it focused solely on transgender women in the capital city and 12 provinces with a high burden of HIV and a large population size of transgender women, limiting its nationwide generalizability. More research are needed to examine the needs of transgender women living in more remote and smaller geographical areas believed to have lower burden of HIV. Aside from the HIV testing, all responses relied on self-report measures that are prone to recall and social desirability response bias.

Conclusions
We found that, despite substantial investments in community-based HIV prevention programs, access to these programs among transgender women in Cambodia remains suboptimal. Only less than half of the women in the study had access to some forms of community-based HIV and legal support services provided by NGOs. To strengthen the response to the needs of this key and vulnerable population, community-based intervention programs should be tailored to respond to the needs of sub-groups of transgender women who are less likely to access the traditional outreach approaches. Using the Respondent Driven Sampling method, our study could reach more than half of participants who had never been reached by any community-based HIV programs. Therefore, a snowball approach could be an effective strategy to reach transgender women who had not been reached by NGOs in the communities.