HIV risk among a sample of drug using male sex workers

This study investigated the characteristics, drug use and sexual risk behaviors of drug-using male sex workers (MSW) in Houston, Texas, USA. A total of 399 MSW had completed a survey of their behavior in the previous 30 days, and reported more than 20 000 sex partners, of whom approximately 2000 were female. Although there were some differences in risk patterns by sexual orientation, heterosexual and bisexual men are not at higher risk because of condom non-use. Male drug users who exchange sex for money are at an extremely high risk of HIV/AIDS and may act as bridges between several at-risk groups [1]. European and American studies have consistently found that male sex workers (MSW) have high rates and frequencies of drug abuse and large numbers of sex partners [2–7]. Most drug using men soliciting sex for money do not self-identify as homosexual [8]. Although sexual self-identification has not been explored as a correlate of HIV risk among MSW, studies involving men who have sex with men suggest that bisexual and heterosexual men may be at a greater risk of HIV infection than gay men [9]. The purpose of this study was to investigate the characteristics, drug use, and sexual risk behaviors of MSW in Houston, Texas, USA. This research was nested in a study of the efficacy of an HIV risk reduction intervention targeting MSW. The sample was recruited between May 1998 and June 2001 using a targeted sampling plan [10]. Locations where men solicit money for sex were identified through interviews with key informants and confirmed by direct observation. Once the locales were identified, a trained outreach worker, with introductions from gatekeepers, approached men and asked if they were interested in being in a health-related study. If responses were positive, men were given a business card and asked to go to a nearby office for screening. After screening and enrollment, men were encouraged to refer other drug using men soliciting money for sex. Participation was open to men who self-identified as male, engaged in street prostitution, were at least 16 years old, reported smoking crack cocaine or injecting an illicit substance in the past 48 h, reported trading sex for money at least once during the previous 7 days, and provided sufficient information to allow a location for follow-up. Once informed consent was obtained, men were asked to provide a urine sample to verify recent cocaine or heroin use. Urine samples were tested using ONTRACK kits. To verify their current involvement in sex work, men were asked questions about soliciting. Responses were compared with current information provided by gatekeepers. A total of 1179 men were screened. Of these, 441 were screened out of the study because of a lack of information needed to schedule follow-up visits, 18 because of a negative urine screen, 233 because sex work could not be verified, and 88 for a negative urine screen and a lack of verifiable sex work. Those screened out were significantly older (33.0 versus 29.9 years, t = 6.52, P < 0.000) and were more likely to be African American (49 versus 28%, χ2 = 51.9, P < 0.000) and heterosexual (26 versus 12%, χ2 = 33.4, P < 0.000). Data were collected using the Sexual Risk Questionnaire in private offices located close to where participants were recruited. The survey took approximately 2 h, with a 10 min break. Participants were paid a small gratuity (US$20). Current drug use was measured by the frequency of using crack cocaine, cocaine, heroin, methamphetamines, and alcohol in the past 30 days. Data on sexual behaviors were collected for vaginal, anal, and oral sex during the past 30 days with main, casual, new, and paying male and female partners. Condom use was measured by asking participants if they used a condom the last time they engaged in each sexual behavior. Committees for the protection of human subjects approved study procedures and the data collection forms. The average age of the 399 MSWs in the sample was 30.3 years (median 30, range 18–53). Half were younger than 30 years (55%), and a quarter were younger than 25 years (23%). Two-thirds were white (65%), the remainder were African American (28%) and Hispanic (6%). More than half (55%) self-identified as bisexual and a third homosexual (33%). Almost all men were single (94%). Two-thirds (64%) did not have a high school education and two-thirds (65%) were unemployed. The remainder worked occasionally or part-time. A tenth (12%) lived in their own home, a third (36%) in someone else's home, a tenth (14%) in hotels, shelters, or flophouses, and the remainder (38%) lived on the streets. Fifty-two men had not been tested for HIV. Reports of HIV infection varied by sexual orientation (χ2 = 51.23, P < 0.000). Thirty-eight per cent of homosexual MSW, 12% of bisexual MSW, and 6% of heterosexual MSW were HIV positive. The sample had high frequencies of drug use, as shown in Table 1. Almost the entire sample smoked crack cocaine (96%), smoking an average of 129.5 times in the 30 days before the interview. About a third injected drugs (38%), injecting from one to two times a week to daily. Of those injecting, 39% reported injecting with needles used by someone else, 36% reported transferring drugs from one syringe to another, and 30% reported sharing paraphernalia. Fourteen per cent of injectors reported being HIV positive. Of those, 52% reported injecting with a syringe used by someone else, 33% reported sharing the drug solution, and 24% reported sharing paraphernalia.Table 1: Drug use and sexual behaviors in 399 male sex workers in Houston, TX, USA.More than half the sample identified as bisexual (55%), one-third as homosexual (33%), and a tenth as heterosexual (12%). MSW reported more than 20 000 sex partners in the previous 30 days. Reports of these sex partners were unique to the individual, but the uniqueness of the partners could not be estimated across MSW. Of the partners reported, almost a tenth were women. A fifth of all sexual encounters involved MSW who were HIV positive. HIV-positive MSW reported 3618 male and 193 female partners. Condom use was inconsistent with all types of partners and varied by sexual orientation and HIV status. Heterosexual men were more likely to use condoms during anal sex with male casual partners than during anal sex with male paying partners. HIV-positive men were less likely to use condoms during vaginal or anal sex with female casual partners and during anal sex with male paying partners. The conclusions that can be drawn from this study are restricted by sampling and measurement. MSW are a ‘hidden and elusive’ population, and it is unlikely that a representative sample of this population is possible. Self-reported data are subject to reporting error [11]. The most likely errors associated with these data are the clustering of events and failure to recall. Whether data are under or overestimates of actual events is unknown. Despite these limitations, the data provide a measure of the level of HIV risk associated with drug use and male sex work. Like other authors [1–7], we found very high frequencies of drug use and sexual risk behaviors. Especially impressive was the number of sex partners. The majority of partners were, as expected, men paying for sex. However, a significant number were women. If we estimate the annual number of sex partners, the number of male partners is approximately 200 000, and female partners 22 000. Even if the total number of unique yearly partners is a fraction of these estimates and only a percentage of encounters involve high-risk behaviors, the figures still represent a substantial level of risk. Reports of condom use by MSW help further to delineate risk levels, and are contrary to findings that heterosexual and bisexual men are at a higher risk than homosexual men. Condom use during anal al and vaginal sex was found to vary by partner type, HIV serostatus and sexual orientation. Although there are some differences in risk patterns by sexual orientation, heterosexual and bisexual men are not at higher risk because of condom non-use. Condom non-use is the rule, not the exception. The reasons for variance in condom use might be personal and relationship factors [12], but for HIV-positive MSW may also include blaming others for infection [13].