Study Design and Sample. Using a cross-sectional study design, we asked TW in Malaysia to answer a survey that evaluated their attitudes towards PrEP and other health-related issues including their ATS use. The eligibility criteria included: 1) 18 years of age or older, 2) assigned male sex at birth and self-identify as female or TW, 3) living in Malaysia, and 4) able to speak or read Bahasa Malaysia, Tamil, or English.
Recruitment. Initially 381 participants were screened for the study, with 7 being ineligible and 13 refusing to participate. As a result, a total of 361 TW completed the study. Participants were recruited via convenience sampling wherein flyers were posted in areas frequented by TW, as well as social media, and referrals from staff at community-based organizations serving TW. Participants were recruited from three states in West Malaysia including Selangor, Penang, and Negeri Sembilan, as well as Kuantan and Pahang in East Malaysia. After providing consent, participants partook in self-administered survey on the Qualtrics software. The survey took approximately 20 minutes to complete. Once completed, the participants were allowed to ask any additional questions regarding the research study and were compensated with 20 Malaysian Ringgit (~ 4 USD) for their time. Full methods are provided in (9).
Study Ethics. This study was approved by the Institutional Review Board of Yale University and the University of Malaysia.
Consent. Participants provided informed consent prior to enrolling in the study.
Measures.
Dependent Variable. The variable of interest, active ATS use, was measured by a combination of two single-item questions, “In the past 30 days, have you used crystal methamphetamine (ice, syabu)?” and “In the past 30 days, have you used ecstasy (E, Pil kuda, MDMA)?”, with a dichotomous response of “yes” or “no” for each.
Participant Characteristics. Variables that described the sociodemographic characteristics of the participants were collected, including age, ethnicity, monthly income, education status (defined as obtained a high school diploma), living in stable housing (defined as residing in a house or apartment alone or with others), religion, relationship status, and mental health status (moderate to severe depressive symptoms). Participant ethnicity was coded per the largest ethnic backgrounds in Malaysia, hence the four categories of Malay, Chinese, Indian, or other. Monthly income was collected as continuous and was coded into three categories to represent the at or below poverty level (≤1169 RM), low income (1170-2208 RM), and middle-upper income (≥2209 RM) ranges per recent income statistics in Malaysia (10). Since Islam is the primary religion in Malaysia, we coded the religion variable to reflect Muslims and those who were not Muslims. The remaining sociodemographic variables mentioned were also dichotomous.
Sexual Health History. The sexual health history variables included recent doctor visit, lifetime STI diagnosis, active hormone therapy use, sexual assault history, physical violence history, and active use of cuci darah. Cuci darah, which when translated from Bahasa Malaysia to English means “blood washing,” and refers to the self-use of unregulated oral, gel capsules that are perceived to contain antibiotics often obtained in the community from peers. All sexual health history questions were measured dichotomously.
Sexual Risk Behavior. Participants were asked questions regarding their sexual risk behavior including: condomless sex with casual partners (defined as a person who is not the primary partner and who does not pay for sex), condomless sex with primary partner (defined as the who does not pay for sex), their sex work status (defined as the frequency of engagement in transactional sex), their involvement in sex work pre-18 years, and their primary method of client solicitation (online vs street-based). All variables were dichotomous except for sex work status which included three categories no sex work, part-time, and full-time statuses.
Criminal Justice Involvement. Variables regarding participants’ criminal history included previous incarcerations, history of drug related arrests, and history of sex work related arrests (defined as having ever been arrested for sex work). All variables were kept as dichotomous.
Statistical Analysis. Of the 381 survey responses, 361 were determined as eligible and therefore were analyzed. A bivariate logistic analysis was conducted on various measures including participant characteristics, sexual health history, and sexual risk behavior variables with the results listed in table 1. Those that were found to be significantly (p<0.1) associated with active ATS use from table 1 were included in the multivariate analysis presented in table 2. Multicollinearity between independent variables was tested using the variance inflation factor (VIF) with the cutoff being at 2.4. Significantly associated (p<0.05) variables from table 2 were then reported and discussed further. The software used for data analyses was IBM SPSS Statistics for Windows version 28.00.