Abstract
Objectives. We studied the HIV risk behaviors of patrons of the 3 commercial sex venues for men in Seattle, Washington.
Methods. We conducted cross-sectional, observational surveys in 2004 and 2006 by use of time–venue cluster sampling with probability proportional to size. Surveys were anonymous and self-reported. We analyzed the 2004 data to identify patron characteristics and predictors of risk behaviors and compared the 2 survey populations.
Results. Fourteen percent of respondents reported a previous HIV-positive test, 14% reported unprotected anal intercourse, and 9% reported unprotected anal intercourse with a partner of unknown or discordant HIV status during the current commercial sex venue visit. By logistic regression, recent unprotected anal intercourse outside of a commercial sex venue was independently associated with unprotected anal intercourse. Sex venue site and patron drug use were strongly associated with unprotected anal intercourse at the crude level. The 2004 and 2006 survey populations did not differ significantly in demographics or behaviors.
Conclusions. Patron and venue-specific characteristics factors may each influence the frequency of HIV risk behaviors in commercial sex venues. Future research should evaluate the effect of structural and individual-level interventions on HIV transmission.
Gay bathhouses and sex clubs—2 key types of commercial sex venues catering exclusively to men—have operated in the United States since at least the first decade of the 20th century.1 With the identification and rapid spread of HIV in the early 1980s among men who have sex with men, these commercial sex venues were quickly thought to be key factors in the epidemic2–4 and were soon the targets of government intervention.5 Yet, despite the central role commercial sex venues were suspected to play in the epidemics of HIV and other sexually transmitted infections in the United States, little epidemiologic research of bathhouses or sex club patrons was done until recently.
Much of the published research involving visitors to commercial sex venues was based on convenience samples.4,6–13 Although several investigators attempted to enroll representative samples of bathhouse patrons, these samples were always limited to a single venue.14–16
Approximately 100 bathhouses and sex clubs were recently found to be in operation within the United States.17 These venues differ in their physical features and amenities. Bathhouses typically offer private rooms for rent and shower facilities and expect that patrons undress and wear only a towel during visits. Sex takes place in private rooms as well as in open areas.18,19 Sex clubs have been described as “a hybrid of bars and bathhouses.”20 Most commonly, sex clubs have no private spaces provided for sex, and patrons typically remain clothed. Sex occurs in open areas, often facilitated by architectural features such as labyrinth-like hallways, small closets, and glory holes (holes in the walls between closets or hallways, primarily used for quasi-anonymous oral sex).19,21
We conducted 2 cross-sectional, observational surveys in 2004 and 2006 of patrons of the 1 sex club and 2 bathhouses located in Seattle, Washington. Our main intent was to describe the populations who visited these venues on any given week during these periods, report the frequency of HIV risk behaviors among these populations, and identify the correlates of high-risk behaviors. A secondary intent was to evaluate the comparability of the 2004 survey population, which achieved a low level of participation (30%, or 373 of 1252 eligible patrons), with the respondents to the much briefer 2006 survey, which enrolled 61% (199 of 20) of eligible patrons. The observed similarity of these populations would assuage concerns about selection bias in the in-depth 2004 survey.
METHODS
The data presented here come from 2 surveys of Seattle patrons of commercial sex venues. The surveys were conducted by the HIV/AIDS Program of Public Health—Seattle & King County in collaboration with researchers from the University of Washington School of Public Health and Community Medicine.
Setting
At the time of the 2004 survey, Seattle was home to 1 sex club and 2 bathhouses (bathhouses A and B). By 2006, the former sex club had added private rooms and shower facilities, but was still primarily used for its sex club–type amenities, operating as a hybrid commercial sex venue. All of the commercial sex venues operated 7 days per week, during daytime hours and throughout each night.
Study Population
We used a time-venue probability-proportional-to-size cluster sampling design22 in each survey to schedule sampling events across club hours of operations. The sampling events were conducted during randomly selected, 4-hour venue, day–time units (VDTUs), with the probability of random selection proportional to the size of the patron population in the specific VDTU. For the 2004 survey, we first based probabilities for selecting each VDTU on estimates of patron volume at times throughout the week observed during a previous study of these 3 venues.7 These probabilities were then revised to reflect observed patron volumes during the first weeks of the survey. The 2006 survey used tallies of patrons observed exiting the venues, maintained by the 2004 survey staff, to establish a probability for each VDTU.
Recruitment and survey completion occurred within the commercial sex venue sites. Patrons were approached for survey recruitment at the end of their visit, just before exiting the facility. A survey staff member recorded the estimated age and race/ethnicity of each patron approached. Patrons who had completed a survey within the respective survey year were deemed ineligible.
Data Collection
Both surveys were anonymous and self-administered. The 2004 survey used audio computer-assisted self-interview technology,23 and the 2006 survey was completed by using pen and paper and a ballot box.
The 2004 survey included an in-depth questionnaire of patron demographic characteristics, history of HIV and sexually transmitted infections, sexual behaviors, drug use, and opinions on proposed HIV/AIDS risk reduction interventions. The questionnaire typically required 30 to 45 minutes to complete, which may have contributed to the observed low participation in this survey, even though we raised the financial remuneration from $10 to $40 for survey participants.
We performed the 2006 study primarily to assess the representativeness of the 2004 sample. Responses were collected by using a brief written questionnaire that included 9 key questions taken verbatim from the 2004 survey. Participants were asked to complete a verbal consent and were reimbursed $5 for their time. We hoped an increased participation rate for this survey would provide some evidence that the sample population responding to the 2004 survey was representative of the overall commercial sex venue population.
Analysis
We first compared responses to the 2006 survey questions with responses to the analogous questions collected as part of the 2004 survey. We also compared the estimated race/ethnicity and age group of the participants and nonparticipants within each survey population. In the main analyses, we further examined the 2004 survey data to describe the characteristics and behaviors of the commercial sex venue population and to identify factors associated with high-risk behavior during the present commercial sex venue visit.
We conducted analyses by using SPSS version 11.5.2 (SPSS Inc, Chicago, IL) and Intercooled STATA 8.1 (Stata Corp, College Station, TX). Unless noted otherwise, simple and multivariate linear and logistic regression models were used for all analyses testing associations, generating odds ratios (ORs) and 95% confidence intervals (CIs). These models were adjusted for clustering within recruitment blocks by using the Huber–White standard error estimation adjustment within STATA. Because of the nonnormal distributions of the continuous variables examined, a rank transformation of each variable was used as the dependent variable in each linear regression model. (In this process, the actual, observed values for the dependent variable were replaced with values reflecting the rank order of the observed value among all participants included in the model.)
We defined any reported unprotected (i.e., without a condom) anal intercourse (UAI) during the current commercial sex venue visit as the main outcome measure in the bivariate and multivariate analyses. Separate questions were asked about unprotected receptive and insertive anal intercourse on each survey. We compared patrons reporting any receptive or insertive UAI with any partner during the current commercial sex venue visit with patrons who reported either no sex or some combination of oral sex or protected anal sex during the visit.
A statistically significant association was defined as evidenced by a P value not exceeding .05. All variables identified in bivariate analyses as statistically associated with the outcome variable or displaying an observable, but not necessarily statistically significant, ordinal trend association with the outcome variable were included in the multivariate logistic regression model.
RESULTS
Comparison of Participants and Nonparticipants
No statistically significant differences in estimated age group (Pearson χ2, P = .7) or estimated race/ethnicity (P = .2) were observed between the participants and nonparticipants in the 2006 study (data not shown). Furthermore, no substantial difference in the racial/ethnic makeup was found in the 2004 study (P = .35). We did find a difference, however, in the estimated age group makeup of the participants and nonparticipants in the 2004 study, with lower proportions of older men agreeing to participate (P < .001). Specifically, 39% of men estimated to be younger than 30 years enrolled in the survey, whereas 33% of men estimated to be between 30 and 39 years, 24% of men perceived to be between 40 and 49 years, and 27% of men estimated to be 50 years or older agreed to participate.
Comparison of the Survey Populations
As shown in Table 1, demographic traits and sexual behaviors reported by the 2004 and 2006 respondents were similar. None of the categorical distributions shown in the table were significantly different between the survey populations. Responses to variables indicating numbers of sex or UAI partners were also examined without categorization, as continuous variables, by using simple linear regression. Again, no statistically significant differences were found between the 2 surveys.
TABLE 1.
Characteristics and Risk Behaviors of the Commercial Sex Venue Survey Populations: Seattle, WA, 2004 and 2006
| 2004 Survey, No. (%) | 2006 Survey, No. (%) | |
| Demographics | ||
| Age, y | ||
| 30–39 | 109 (29.4) | 58 (29.6) |
| 40–49 | 145 (39.1) | 67 (34.2) |
| ≥ 50 | 60 (22.9) | 39 (28.1) |
| Race/ethnicity | ||
| White | 262 (70.4) | 142 (72.1) |
| Asian/Pacific Islander | 33 (8.9) | 13 (6.6) |
| Latino | 32 (8.6) | 18 (9.1) |
| Black | 21 (5.6) | 19 (9.6) |
| Other/mixed | 24 (6.5) | 5 (2.5) |
| Resident of King County | ||
| Yes | 269 (74.7) | 143 (72.2) |
| No | 91 (25.3) | 55 (27.8) |
| Education | ||
| High school or less | 9 (2.4) | |
| Some college | 159 (42.6) | |
| College degree | 115 (30.8) | |
| Graduate degree | 90 (24.1) | |
| Sexual identification | ||
| Gay | 281 (78.5) | |
| Bisexual | 58 (16.2) | |
| Other | 19 (5.3) | |
| Has domestic partner or spouse | ||
| Yes | 91 (25.4) | |
| No | 267 (74.6) | |
| Behaviors during current visit | ||
| Oral sex | ||
| HIV+ patrons | 42 (84.0) | |
| All other patrons | 233 (76.6) | |
| Anal sex | ||
| HIV+ patrons | 16 (32.0) | |
| All other patrons | 110 (36.2) | |
| Unprotected anal intercourse | ||
| HIV+ patrons | 9 (18.0) | |
| All other patrons | 40 (13.2) | |
| Unprotected anal intercourse with UDS partner | ||
| HIV+ patrons | 6 (12.0) | |
| All other patrons | 25 (8.3) | |
| Total no. of sex partners | ||
| 0 | 60 (16.1) | 24 (12.2) |
| 1 | 98 (26.3) | 54 (27.4) |
| 2 | 76 (20.4) | 44 (22.3) |
| ≥ 3 | 139 (37.3) | 75 (38.1) |
| Total no. of insertive UAI partners | ||
| 0 | 334 (90.0) | 174 (88.3) |
| 1 | 27 (7.3) | 15 (7.6) |
| 2 | 6 (1.6) | 8 (4.1) |
| ≥ 3 | 4 (1.1) | 0 (0.0) |
| Total no. of receptive UAI partners | ||
| 0 | 347 (93.8) | 181 (91.4) |
| 1 | 16 (4.3) | 12 (6.1) |
| 2 | 5 (1.4) | 4 (2.0) |
| ≥ 3 | 2 (0.5) | 1 (0.5) |
| Told HIV status to all unprotected anal intercourse partners | ||
| Yes | 22 (44.0) | |
| No | 28 (56.0) | |
| Substances used in previous 24 h | ||
| Alcohol | 115 (30.8) | |
| Nitrite inhalants | 94 (25.2) | |
| Erectile dysfunction medication | 50 (13.4) | |
| Methamphetamine | 40 (10.7) | |
| Marijuana | 39 (10.5) | |
| Cocaine | 12 (3.2) | |
| Any druga | 153 (41.7) | |
| Behaviors, past 3 mo | ||
| Total unprotected anal intercourse partners in all commercial sex venues, no. | ||
| 0 | 261 (72.5) | 149 (76.0) |
| 1 | 40 (11.1) | 14 (7.1) |
| 2 | 21 (5.8) | 9 (4.6) |
| 3–5 | 21 (5.8) | 6 (3.1) |
| ≥ 6 | 17 (4.7) | 18 (9.2) |
| Total unprotected anal intercourse partners outside ofcommercial sex venues no. | ||
| 0 | 277 (77.2) | 135 (69.6) |
| 1 | 41 (11.4) | 26 (13.4) |
| 2 | 17 (4.7) | 14 (7.2) |
| 3–5 | 13 (3.6) | 13 (6.7) |
| ≥ 6 | 11 (3.1) | 6 (3.1) |
| No. of visits | ||
| 1 | 78 (21.7) | |
| 2–4 | 115 (31.9) | |
| 5–10 | 96 (26.7) | |
| 11–20 | 50 (13.9) | |
| ≥ 20 | 21 (5.8) | |
| Had sex at > 1 commercial sex venue | ||
| Yes | 170 ( 47.2) | |
| No | 190 (52.8) | |
| Unprotected anal intercourse with UDS partner in a commercial sex venue | ||
| Yes | 82 (22.8) | |
| No | 279 (77.2) | |
| Unprotected anal intercourse with UDS partner outside of a commercial sex venue | ||
| Yes | 39 (10.9) | |
| No | 319 (89.1) | |
| Behaviors, past y | ||
| No. of anal or vaginal sex partners | ||
| None | 68 (19.2) | |
| 1 | 45 (12.7) | |
| 2–5 | 103 (29.1) | |
| 6–10 | 37 (10.5) | |
| 11–20 | 42 (11.9) | |
| > 20 | 59 (16.7) | |
| Vaginal or anal sex with a woman | ||
| Yes | 51 (14.4) | 29 (14.8) |
| No | 304 (85.6) | 167 (85.2) |
| HIV/STI-related behaviors | ||
| Self-reported HIV status | ||
| HIV- | 284 (80.2) | |
| HIV+ | 50 (14.1) | |
| Never tested | 20 (5.6) | |
| Last HIV testb | ||
| Past 3 mo | 126 (35.3) | |
| 3–6 mo | 73 (20.4) | |
| 6–12 mo | 54 (15.1) | |
| ≥ 1 y | 84 (23.5) | |
| Never | 20 (5.6) | |
| STI diagnosis (past y) | ||
| Gonorrhea | 18 (5.0) | |
| Chlamydia | 16 (4.5) | |
| Chlamydia and gonorrhea | 3 (0.8) | |
| Other STI | 17 (4.8) | |
| Tested, no STI | 182 (51.0) | |
| Not tested | 121 (33.9) | |
Note. UAI = unprotected anal intercourse; UDS = unknown or discordant HIV status; STI = sexually transmitted infection. No differences between 2004 and 2006 were significant at P ≤ .05. All tests of significance were performed using logistic regression with standard errors adjusted for clustering within venue, day–time unit. The 2006 survey included only 9 questions from the 2004 survey.
Excluding marijuana and alcohol.
HIV-infected participants were excluded.
Description of the 2004 Survey Population
The remaining analyses discussed are limited to data from the 2004 survey. As shown in Table 1, 14% (50 of 354) of these respondents reported a previous HIV-positive diagnosis. Of the remaining men, most (56%) had tested HIV-negative within the past 6 months, and another 15% had tested HIV-negative in the previous 6 to 12 months. Although most of the men reported a recent HIV test, 37% reported having had UAI since this last test.
Not shown in the table is the frequency of any UAI outside a commercial sex venue in the past 3 months, stratified by commercial sex venue sites visited over the same period. Among the 74 respondents who visited only the sex club, 12% reported UAI; 25% of the 114 visitors of 1 of the 2 bathhouses reported UAI; and 26% of 169 men who reported visiting multiple commercial sex venues reported this behavior (P < .05 for each compared with the sex club).
Respondents were asked to report the HIV status, if known, of each reported UAI partner during the current commercial sex venue visit and over the previous 3 months. Therefore, we were able to determine whether UAI partners were of unknown or discordant HIV status. Less than one tenth of all men reported a UAI partner of unknown or discordant HIV status during the current visit, although just less than 23% engaged in UAI with such a partner in a commercial sex venue during the previous 3 months. Fewer men (11%) reported UAI with a partner of unknown or discordant HIV status outside a commercial sex venue in the 3 months before the survey.
Alcohol was the most commonly reported substance used in the previous 24 hours (31%), followed by nitrite inhalants, which were used by 25% of the respondents. In all, 42% of the men reported using methamphetamine, nitrite inhalants or inhalants, ecstasy, cocaine, GHB (gamma-hydroxybutyric acid), ketamine, or medication for erectile dysfunction in the previous 24 hours.
Correlates of Unprotected Anal Intercourse in a Commercial Sex Venue
Results of the bivariate analyses examining relations between various factors and UAI during the current commercial sex venue visit are shown in Table 2. Relatively few men in the sex club reported UAI (P < .005 for the difference between each bathhouse and the sex club). A large disparity in the proportion of men who reported recent UAI outside a commercial sex venue was observed among patrons who reported UAI (55%) and did not report UAI (18%) during the current commercial sex venue visit (P < .001). Similarly, a higher proportion of men who reported UAI also reported drug use in the previous 24 hours (P < .001).
TABLE 2.
Crude Relationships Between Selected Factors and Unprotected Anal Intercourse During Current Visit of Commercial Sex Venue Patrons: Seattle, WA, 2004
| Unprotected Anal Sex During Current Visit |
||
| Yes, no. (%) | No, no. (%) | |
| Site* | ||
| Sex club | 9 (17.0) | 130 (40.8) |
| Bathhouse A | 28 (52.8) | 127 (39.8) |
| Bathhouse B | 16 (30.2) | 62 (19.4) |
| Any unprotected anal sex outside a commercial sex venue in past 3 mo** | ||
| Yes | 27 (55.1) | 55 (17.6) |
| No | 22 (44.9) | 257 (82.4) |
| Any drug use todaya,** | ||
| Yes | 34 (68.0) | 119 (37.5) |
| No | 16 (32.0) | 198 (62.5) |
| No. of men had sex with at all commercial sex venues in past 3 mo | ||
| 0–2 | 9 (19.1) | 84 (27.2) |
| 3–5 | 8 (17.0) | 60 (19.4) |
| 6–13 | 14 (29.8) | 94 (30.4) |
| ≥ 14 | 16 (34.0) | 71 (23.0) |
| Met a partner on the Internet, past 3 mo | ||
| Yes | 16 (32.7) | 69 (22.1) |
| No | 33 (67.3) | 243 (77.9) |
| Previous HIV+ test | ||
| Yes | 9 (18.4) | 41 (13.4) |
| No | 40 (81.6) | 264 (86.6) |
| Told HIV status to all anal sex partnersb | ||
| Yes | 6 (27.3) | 29 (36.7) |
| No | 16 (72.7) | 50 (63.3) |
| Visited another bathhouse/sex club in past 3 mo | ||
| Yes | 24 (51.1) | 146 (46.6) |
| No | 23 (48.9) | 167 (53.4) |
| Currently have domestic partner or spouse | ||
| Yes | 14 (28.6) | 77 (24.9) |
| No | 35 (71.4) | 232 (75.1) |
| More than 1 commercial sex venue visit in past 3 mo | ||
| Yes | 11 (23.4) | 67 (21.4) |
| No | 36 (76.6) | 246 (78.6) |
| Oral or anal sex in a private roomc | ||
| Yes | 39 (88.6) | 102 (75.0) |
| No | 5 (11.4) | 34 (25.0) |
Note. All tests of significance were performed using logistic regression with standard errors adjusted for clustering within venue, day–time unit.
Excluding marijuana and alcohol
Limited to men who reported anal sex during the current visit.
Limited to bathhouse A and B patrons who reported any oral or anal sex.
*P < .005; **P < .001.
The number of sex partners in the commercial sex venues over the previous 3 months was not significantly associated with UAI during the current visit. However, as the category representing the number of sex partners at the commercial sex venue over the previous 3 months increased, in general so did the proportion of men reporting UAI during the current visit. The other factors presented in Table 2 had no additional, statistically significant relations with UAI during the current visit.
The frequency of key risk behaviors among patrons of the 3 venues is summarized in Figure 1. Each behavior, except for nitrite inhalants use, was least common at the sex club, and most differences were statistically significant.
FIGURE 1.
Frequency of sexual risks and drug use, by commercial sex venue site: Seattle, WA, 2004.
Note. UDS = unknown or discordant HIV status. Nitrite inhalants are nitrate inhalants. Sexual behaviors during current commercial sex venue visit; drug use in previous 24 hours. Also, statistically significant crude associations were observed between both crystal methamphetamine use and UAI, and nitrite inhalants use and UAI.
aStatistically significant (P < .05) difference between sex club and bathhouse A.
bStatistically significant (P < .05) difference between sex club and bathhouse B.
cStatistically significant (P < .05) difference between bathhouse A and bathhouse B.
The results of the multivariate logistic regression model are presented in Table 3. Only 1 variable, recent UAI outside a commercial sex venue (OR = 4.28; 95% CI = 2.06, 8.91), was associated with UAI during the current visit. Specific commercial sex venue site, drug use, and the number of sex partners in a commercial sex venue in the past 3 months was not significantly associated with UAI, although the ORs for the bathhouses (versus the sex club) were equal to 3.1 (95% CI = 0.93, 10.48) for bathhouse A and 2.9 (95% CI = 0.77, 11.07) for bathhouse B. No interaction terms were statistically significant.
TABLE 3.
Results of Multivariate Logistic Regression Model Predicting Unprotected Anal Intercourse During the Current CSV Visit, Seattle, WA, 2004
| Main Model: Entire Survey Population (N = 353), AOR (95% CI) | Exploratory Model: Bathhouse Patrons Who Reported Oral or Anal Sex (n = 171), AOR (95% CI) | |
| Site (vs sex club)a | ||
| Bathhouse A | 3.12 (0.93, 10.48) | … |
| Bathhouse B | 2.93 (0.77, 11.07) | 0.99 (0.40, 2.49) |
| Any unprotected anal sex outside a commercial sex venue in past 3 mo | 4.28 (2.06, 8.91) | 5.58 (2.16, 14.40) |
| Any drug useb | 2.05 (0.93, 4.50) | 2.47 (0.86, 7.15) |
| No. of bathhouse sex partners in past 3 mo (vs 2 or fewer) | ||
| 3–5 | 1.16 (0.42, 3.21) | 1.03 (0.29, 3.65) |
| 6–13 | 1.55 (0.60, 4.03) | 1.08 (0.37, 3.20) |
| ≥ 14 | 1.99 (0.74, 5.38) | 1.25 (0.39, 4.02) |
| Oral or anal sex in a private room during current visit | … | 3.05 (1.03, 9.01) |
Note. AOR = adjusted odds ratio; CI = confidence interval. Ellipses indicate that data were not applicable. All tests of significance were performed by using logistic regression with standard errors adjusted for clustering within venue, day–time unit.
Bathhouse A referent in exploratory model.
In the past 24 hours; excludes marijuana and alcohol.
Subanalysis of Bathhouse Patrons
We conducted a subanalysis of patrons of bathhouses A and B who reported any oral or anal sex during the current commercial sex venue visit to explore the relations between use of the private rooms and UAI. Only 22% of these 180 sexually active patrons did not report oral or anal sex in a private room during the visit (Table 2). Of these men, 13% reported UAI; among the remaining men who had, in fact, used a private room for sex, 28% reported UAI (P = .04). Inclusion of this factor in a multivariate model yielded an OR of 3.05 (95% CI = 1.03, 9.01) for the variable indicating sex in a private room (Table 3). Sex outside a commercial sex venue remained strongly associated with UAI (OR = 5.58; 95% CI = 2.16, 14.40). Other variables in this model were not significantly correlated with UAI.
DISCUSSION
Our 2004 sample of patrons of commercial sex venues showed that high-risk behaviors for HIV in commercial sex venues were independently related to a respondent's likelihood of risk-taking elsewhere. The broader analysis suggested that specific characteristics of the commercial sex venue site and patron drug use may also be related to UAI. Furthermore, we found that the having UAI was relatively uncommon (14%) during any particular commercial sex venue visit, although it was somewhat more common (24%) when considering all of an individual's commercial sex venue visits over a 3-month period. Importantly, the self-reported prevalence of HIV infection in this population was 14%, which is nearly identical to that observed in a recent population-based sample of Seattle men who have sex with men.24
These findings are of particular interest because they reflect commercial sex venue data collected by the use of probability sampling methods. Also, they provide the first profile of a population across all gay bathhouses and sex clubs within a US municipality.
Limitations
Data from our surveys, however, are limited in their ability to identify causal relations because of the cross-sectional, observational survey designs used. Also, if our source population were intended to include all men visiting a commercial sex venue over a longer period such as a year, the venue-based sampling approach would have likely resulted in a bias toward the inclusion of frequent commercial sex venue visitors. The stratified, weighted random selection of sampling events was designed to give all commercial sex venue patron visits across each week an equal probability of inclusion in the survey. However, the frequency of visits by an individual patron could not be accounted for in this design, and so men visiting most often during survey recruitment would have the highest probability of survey enrollment. Therefore, our sample may reflect the Seattle commercial sex venue population over a short period such as a day or a week but may overrepresent frequent visitors over a longer duration such as a year. Future research of this type may consider adopting methods that allow for adjustments based on patron visit frequency.25
Previous research found that frequent attendance at gay venues may be associated with higher levels of HIV risk behaviors.26 As shown in Table 2, we did not observe a difference in the probability of UAI during the current commercial sex venue visit between frequent and infrequent visitors. Therefore, our bivariate and multivariate findings may not have been affected by this issue. Some of the descriptive findings, however, particularly those presenting behaviors over 3 months or a year, were more likely influenced by patrons’ frequency of commercial sex venue attendance.
The sample size of the 2006 survey was designed to have adequate statistical power to detect meaningful differences in key variables, as compared with the 2004 values. For the main analyses presented here, however, our findings are limited by the relatively small sample size for the 2004 survey. It is possible that some associations did not appear to be statistically significant in our analysis because of an inadequate sample size.
Much of the information solicited from survey respondents was sensitive and potentially stigmatizing, thereby raising the possibility of reporting bias.27 We attempted to minimize reporting bias through the use of anonymous, self-administered surveys, which have been shown to increase the frequency of reporting of such private, stigmatizing information.28–31
Our main analyses were based on a 2004 survey with a low, 30% participation rate. Recruiters during this survey perceived that younger men were somewhat more likely to enroll in the survey. These factors raise concerns about nonparticipation bias. Using comparisons of key demographic and behavioral indicators, however, our analyses showed that the respondents to the 2004 survey did not appear to be substantially different from the 2006 survey population. The latter population was designed to be representative of Seattle commercial sex venue patrons on any given day during administration of the survey. Yet, the proportion of eligible patrons who chose to complete the 2006 survey was only 61%, which was lower than we wished to achieve to minimize potential selection bias. Therefore, we cannot rule out the possibility that the patrons who chose to complete our surveys were in some important ways not representative of all Seattle commercial sex venue patrons during survey recruitment.
Conclusions
Only 1 comparable study of a commercial sex venue population conducted since the widespread introduction of highly active antiretroviral therapy has been described in the literature. In that study, the investigators used similar probability sampling methods in a 2001 survey of patrons of a bathhouse located in the western United States.16 Their study found a slightly lower frequency of UAI during individual commercial sex venue patron visits (11%) and during the previous 3-month period in commercial sex venues and public cruising areas (21%). Similar levels of substance use and of previous HIV-positive tests were also reported in this study. Each of the independent predictors of UAI identified in our analysis was also found to be a statistically significant correlate, although the authors also identified receiving an HIV-positive diagnosis, previous UAI in a public sex venue, attendance at multiple commercial sex venues, and having a domestic partner or spouse as important correlates of UAI.
The lower frequency of UAI among patrons of the sex club, as compared with the 2 bathhouses, could be attributed to features distinguishing the bathhouses. These included (1) private rooms with horizontal surfaces large enough for an adult to comfortably recline on, (2) shower facilities and the routine provision of towels, and (3) the removal of clothes during typical patron visits.
The relation between private rooms and risks in bathhouses has long been the subject of debate.32 Less public discussion has taken place on the provision of beds and other horizontal surfaces large enough for reclining, yet one could reasonably posit that these features accommodate anal sex. Similarly, many men prefer to wash themselves after anal sex, so providing towels and showers may increase the probability that some patrons will have anal sex. Finally, removing clothing clearly makes it easier for patrons to have sex, especially anal sex.
Our examination of the relationship between use of private rooms for sex and UAI did show a statistically significant association between these variables. As we noted, the proportion of men who reported UAI during the current visit was more than twice as high among private room users as among patrons who did not have sex in a private room. This provides some evidence that the availability of private rooms in commercial sex venues may increase the frequency of UAI.
We cannot assume that the removal of private rooms would not give rise to other, perhaps difficult to foresee, consequences. This concern is perhaps supported by our survey data, which show that the commercial sex venues in which UAI occurs most often also seem to attract men who are more likely to have UAI partners outside of commercial sex venues and men who use crystal methamphetamine. This cohort of men who seek out risky sex and drug use may have chosen bathhouses over the sex club specifically because they offer more opportunities for risky sex; altering these spaces could displace a segment of these high-risk men to venues out of the reach of public health efforts.
The limitations noted above, in conjunction with the paucity of comparable epidemiologic research on the subject of commercial sex venues, suggest that our findings should be considered to be exploratory in nature. More research is needed to elucidate the effects of individual and venue characteristics on HIV risk behaviors and, further, to identify the indirect effects of changes in these structural characteristics on risky sex and transmission of HIV and other sexually transmitted infections in the larger communities surrounding commercial sex venues.
Experimental research on commercial sex venue structural factors may prove to be logistically difficult to conduct. Future efforts might therefore include further observational studies to better elucidate the patterns of HIV risk behaviors by sex club and bathhouse patrons, both inside commercial sex venues and elsewhere, and exploratory mathematical models of HIV transmission among commercial sex venue patrons under various structural intervention scenarios. Sexual risk behaviors outside commercial sex venues reported by this population may provide clues about the nature of displaced sexual partnerships if, indeed, some patrons were to reduce their use of commercial sex venues after structural interventions.
Acknowledgments
The 2004 and 2006 surveys were funded by the Seattle–King County Public Health HIV/AIDS Program. William Reidy was supported by funding from the Seattle–King County Public Health HIV/AIDS Program and the State of Washington through the Department of Epidemiology, University of Washington School of Public Health and Community Medicine.
The authors recognize the following individuals for their contributions to the surveys and to this article: Frank Chaffee, Michael Underhill, James Hughes, Jsani Henry, Jef Saint DeLore, Joe Tinsley, Thomas Koepsell, and Brent Zimmerman.
Human Participant Protection
Human participants approval for the 2006 study was provided by the University of Washington institutional review board. The 2004 effort was conducted as a public health needs assessment and therefore no institutional review board approval was sought before data collection; however, the University of Washington institutional review board retrospectively approved the use of these data for research purposes in 2007. (Both approvals were contained within application number 05-7703-E/G 02.)
References
- 1.Tewksbury R. Bathhouse intercourse: structural and behavioral aspects of an erotic oasis. Deviant Behav. 2002;23(1):75–112 [Google Scholar]
- 2.Darrow WW, Jaffe HW, Curran JW. Passive anal intercourse as a risk factor for AIDS in homosexual men. Lancet. 1983;322(8342):160. [DOI] [PubMed] [Google Scholar]
- 3.Jaffe HW, Choi K, Thomas PA, et al. National case-control study of Kaposi's sarcoma and Pneumocystis carinii pneumonia in homosexual men: part 1. Epidemiologic results. Ann Intern Med. 1983;99(2):145–151 [DOI] [PubMed] [Google Scholar]
- 4.McKusick L, Horstman W, Coates TJ. AIDS and sexual behavior reported by gay men in San Francisco. Am J Public Health. 1985;75(5):493–496 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Burris S. Legal aspects of regulating bathhouses: cases from 1984 to 1995. J Homosex. 2003;44(3/4):131–151 [DOI] [PubMed] [Google Scholar]
- 6.Peterson JL, Coates TJ, Catania JA, Middleton L, Hilliard B, Hearst N. High-risk sexual behavior and condom use among gay and bisexual African-American men. Am J Public Health. 1992;82(11):1490–1494 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Spielberg F, Branson BM, Goldbaum GM, Kurth A, Wood RW. Designing an HIV counseling and testing program for bathhouses: the Seattle experience with strategies to improve acceptability. J Homosex. 2003;44(3/4):203–220 [DOI] [PubMed] [Google Scholar]
- 8.Wolf FC, Judson FN. Intensive screening for gonorrhea, syphilis, and hepatitis B in a gay bathhouse does not lower the prevalence infection. Sex Transm Dis. 1980;7(2):49–52 [DOI] [PubMed] [Google Scholar]
- 9.Merino HI, Judson FN, Bennett D, Schaffnit TR. Screening for gonorrhea and syphilis in gay bathhouses in Denver and Los Angeles. Public Health Rep. 1979;94(4):376–379 [PMC free article] [PubMed] [Google Scholar]
- 10.Judson FN, Miller KG, Schaffnit TR. Screening for gonorrhea and syphilis in the gay baths—Denver, Colorado. Am J Public Health. 1977;67(8):740–742 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Woods WJ, Sabatino J, Bauer PL, Adler B, Dilley JW, Binson D. HIV testing in gay sex clubs. Int J STD AIDS. 2000;11(3):173–175 [DOI] [PubMed] [Google Scholar]
- 12.Huebner DM, Binson D, Woods WJ, Dilworth SE, Neilands TB, Grinstead O. Bathhouse-based voluntary counseling and testing is feasible and shows preliminary evidence of effectiveness. J Acquir Immune Defic Syndr. 2006;43(2):239–246 [DOI] [PubMed] [Google Scholar]
- 13.Spielberg F, Branson BM, Goldbaum GM, et al. Choosing HIV counseling and testing strategies for outreach settings: a randomized trial. J Acquir Immune Defic Syndr. 2005;38(3):348–355 [PubMed] [Google Scholar]
- 14.Van Beneden CA, O'Brien K, Modesitt S, Yusem S, Rose A, Fleming D. Sexual behaviors in an urban bathhouse 15 years into the HIV epidemic. J Acquir Immune Defic Syndr. 2002;30(5):522–526 [DOI] [PubMed] [Google Scholar]
- 15.Richwald GA, Morisky DE, Kyle GR, Kristal AR, Gerber MM, Friedland JM. Sexual activities in bathhouses in Los Angeles County: implications for AIDS prevention education. J Sex Res. 1988;25(2):169–180 [Google Scholar]
- 16.Woods WJ, Binson D, Blair J, Han L, Spielberg F, Pollack LM. Probability sample estimates of bathhouse sexual risk behavior. J Acquir Immune Defic Syndr. 2007;45(2):231–238 [DOI] [PubMed] [Google Scholar]
- 17.Woods WJ, Tracy D, Binson D. Number and distribution of gay bathhouses in the United States and Canada. In: Woods WJ, Binson D, eds. Gay Bathhouses and Public Health Policy. Binghamton, NY: Harrington Park Press; 2003:55–70 [DOI] [PubMed] [Google Scholar]
- 18.Hoffman M. The Gay World: Male Homosexuality and the Social Creation of Evil. New York, NY: Basic Books; 1968 [Google Scholar]
- 19.Woods WJ, Binson D, Mayne TJ, Gore LR, Rebchook GM. Facilities and HIV prevention in bathhouse and sex club environments. J Sex Res. 2001;38(1):68–74 [DOI] [PubMed] [Google Scholar]
- 20.Lindell J. Public Space For Public Sex. In: Colter EG, Hoffman W, Pendleton E, Redick A, Serlin D, eds. Policing Public Sex. Boston, MA: South End Press; 1996:73–80 [Google Scholar]
- 21.McCoy CB, Inciardi JA. Sex, Drugs, And the Continuing Spread of AIDS. Los Angeles, CA: Roxbury; 1995 [Google Scholar]
- 22.Muhib FB, Lin LS, Stueve A, et al. A venue-based method for sampling hard-to-reach populations. Public Health Rep. 2001;116(suppl 1):216–222 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Mackenzie SL, Kurth AE, Spielberg F, et al. Patient and staff perspectives on the use of a computer counseling tool for HIV and sexually transmitted infection risk reduction. J Adolesc Health. 2007;40(6):572.e9–572.e16 [DOI] [PubMed] [Google Scholar]
- 24.Brewer DD, Golden MR, Handsfield HH. Unsafe sexual behavior and correlates of risk in a probability sample of men who have sex with men in the era of highly active antiretroviral therapy. Sex Transm Dis. 2006;33(4):250–255 [DOI] [PubMed] [Google Scholar]
- 25.Shepard DS, Neutra R. A pitfall in sampling medical visits. Am J Public Health. 1977;67(8):743–750 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Xia Q, Tholandi M, Osmond DH, et al. The effect of venue sampling on estimates of HIV prevalence and sexual risk behaviors in men who have sex with men. Sex Transm Dis. 2006;33(9):545–550 [DOI] [PubMed] [Google Scholar]
- 27.Catania JA. A framework for conceptualizing reporting bias and its antecedents in interviews assessing human sexuality. J Sex Res. 1999;36(1):25–38 [Google Scholar]
- 28.Kissinger P, Rice J, Farley T, et al. Application of computer-assisted interviews to sexual behavior research. Am J Epidemiol. 1999;149(10):950–954 [DOI] [PubMed] [Google Scholar]
- 29.Newman JC, Des J, Turner CF, Gribble J, Cooley P, Paone D. The differential effects of face-to-face and computer interview modes. Am J Public Health. 2002;92(2):294–297 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Kurth AE, Martin DP, Golden MR, et al. A comparison between audio computer-assisted self-interviews and clinician interviews for obtaining the sexual history. Sex Transm Dis. 2004;31(12):719–726 [DOI] [PubMed] [Google Scholar]
- 31.Durant LE, Carey MP. Self-administered questionnaires versus face-to-face interviews in assessing sexual behavior in young women. Arch Sex Behav. 2000;29(4):309–322 [DOI] [PubMed] [Google Scholar]
- 32.Disman C. The San Francisco bathhouse battles of 1984: civil liberties, AIDS risk, and shifts in health policy. J Homosex. 2003;44(3/4):71–129 [DOI] [PubMed] [Google Scholar]