Woods, William J PhD*; Binson, Diane PhD*; Blair, Johnny BA†; Han, Lei PhD*; Spielberg, Freya MD‡; Pollack, Lance M PhD*
Associating HIV transmission with gay bathhouses is as old as the epidemic itself.1,2 Results of analyses from the Urban Men's Health Study confirmed that the relation is real only insofar as men who engage in HIV sexual risk behavior are significantly more likely to go to bathhouses than men who do not engage in risk behavior.3 Despite the common assumption that bathhouses facilitate this high-risk sex, where risk behavior occurs remains an empiric question.
A few studies of sex in the bathhouses have been done but none are adequately able to describe the prevalence of risk behavior that actually takes place in these venues.4 These studies fall into 4 broad categories. The first of these are studies that report data collected through direct inspection.5-8 Reports through inspection are problematic in that observers cannot assess what goes on behind closed doors and cannot be everywhere at once, even when several people go to observe. A similar design uses qualitative interviews with nonrandomly selected patrons9,10 who serve as expert witnesses of bathhouse behavior. One important study in the 1970s used a combination of these first 2 approaches.11 Another type of bathhouse study relies on patron surveys, but these were done with convenience samples and, as such, produce unreliable estimates of risk.12-14
A final data source is studies of men in HIV epicenters, with specific questions about going to bathhouses or sex clubs.3,15-19 Only 1 study used a probability sampling method.3 This study examined the frequency of bathhouse visits and risk behavior but did not differentiate as to whether the sexual risk behavior occurred at bathhouses or other places where men have sex. Nevertheless, data from this study suggest that bathhouses are places that attract the highest risk men who have sex with men (MSM), specifically men who visit multiple types of public sex environments. Another study examined the various ways different cities regulate behavior in gay bathhouses and found that the regulations had no effect on the overall prevalence of risk behavior among adult MSM who go to bathhouses.20
The present study was conducted during the last 5 weeks of 2001 at a bathhouse located in the western United States. This venue is a typical North American gay bathhouse, perhaps among the larger ones in terms of space and the number of patrons it can accommodate at a single time.6,8,21 As is typical of these establishments, the business is privately owned and operated, caters exclusively to men, and is open 24 hours a day 7 days a week. The venue provided a full range of spaces typical in a gay bathhouse, including a steam room, a weight room, showers, a whirlpool, an open play area designed as a maze, and private rooms for rent. Two of the main ways in which clubs differ is whether they permit sexual activity in open areas of the club and whether they have private rooms, although these physical differences do not seem to make a difference in the risk behavior of those who patronize the venues.20 Clients pay a fee to enter for up to 6 hours, and at the end of that time, they can pay another entry fee for 1 (and only 1) more 6-hour period. Once inside the club, the only clothing worn is generally a towel provided by the club.
The survey data make possible the first description of a population of bathhouse patrons and their HIV risk behavior while at the bathhouse. The study also examines sexual risk in the 3 months before the visit and whether prior risk, along with client characteristics, is a correlate of predictors of men's HIV risk behavior inside the club.
The participants were a sample of men leaving a gay bathhouse (n = 400). A sampling statistician developed a probability sampling plan based on patron exit data provided by the bathhouse. A 2-stage cluster sample design was used. In the first stage, the 24-hour day was divided into 3-hour shifts (recruitment shifts) and shifts were selected; in the second stage, respondents visiting the establishment during those shifts were sampled. The shifts were selected with probabilities proportional to the expected number of visitors. A fixed cluster size of exiting patrons in each sampled shift was then chosen with equal probabilities. The 2 stages taken together produce an equal probability sample of visits. Thus, during each recruitment shift, men were selected based on a random start number and a designated sampling fraction or interval (ie, every “nth” man exiting was selected for the sample). It is important to note that the sample is of visits rather than visitors. The probability of an individual visitor is proportional to the number of times he visited the establishment during the data collection period. As a part of the estimation procedure, the sample of visits was weighted to convert it to a sample of individual visitors.
Data collection occurred on the last 36 consecutive days of 2001; this period included 288 3-hour work shifts. From these 288 shifts, 97 (33.7%) were randomly selected for data collection. Between 1 and 4 shifts were selected for data collection for each of the 36 days. Fifteen shifts were selected for data collection during the first week from Monday through Sunday; 18 in the second week; and 20 in each of the third, fourth, and fifth weeks (plus 4 shifts on the “extra” 36th day).
To produce unbiased sample estimates, weights were developed to reflect probability of selection, multiple visits during the data collection period, nonresponse, and noncoverage. All data in this report are weighted. Details of the sampling design and selection of procedures and weights are described in more detail elsewhere.22
Every man exiting the bathhouse during selected shifts was counted. Each man selected for inclusion was approached and recruited to do an interview and was considered eligible unless he had already participated or was unable to give consent because of insufficient knowledge of English or cognitive impairment.
Three-man teams were assigned to each randomly selected shift; 1 man served as counter, and the other 2 were available for recruiting and interviewing each “nth” man checking out of the club. After recruitment, men were taken to a private room in the club and, after giving verbal consent to participate, completed an interviewer-administered questionnaire. Participation was voluntary, anonymous, and reimbursed. All procedures and measures were reviewed and approved by the Institutional Review Board of the University of California, San Francisco.
Data were collected using an exit survey. The interviewer recorded the interview start time and then asked the respondent what time he had arrived at the bathhouse on that particular visit. Before administering the survey, the interviewer defined sex for the purposes of the study. The respondent was told that sexual behavior questions would be specifically about anal and oral sex only and that he should consider any penis-to-anus or penis-to-mouth contact as “sex,” no matter how brief or long it lasted and whether or not he or his partner reached orgasm. This definition was repeated throughout the course of the interview. After first discussing this definition, the respondent was asked the number of men he had “sex” with during that visit to the bathhouse. Subsequent questions asked about the number of anal sex partners, specifically about receptive and insertive anal intercourse and whether or not it included use of condoms or resulted in ejaculation before withdrawal. (For this analysis, the number of anal sex partners was calculated by summing the number of receptive and insertive partners.) Similar questions were then asked about oral sex. Next, the respondent was asked about group sex. Group sex was defined as 3 or more men together when the survey participant was engaged in any sex (as previously defined) with at least 1 of the other men involved. If a third person simply watched, it was not considered group sex. We then asked him to name any recreational drugs (including alcohol) that he might have taken in the past 24 hours. For each substance he reported, we asked him if he were still feeling the effects of the drug during the current bathhouse visit.
The next section of the survey asked about past sexual behavior and visits to the bathhouse and other sex venues. Among these items, we asked about where he had sex in the past 3 months, specifically, his own or someone else's home, a bathhouse/sex club, a cruising area (eg, park, beach, adult bookstore), and other places he specified. For each place he had sex, we followed up with 2 questions: whether he had insertive anal sex without a condom and whether he had receptive anal sex without a condom.
This was followed by questions about testing for HIV and other sexually transmitted infections (STIs). In the concluding section, the respondent provided basic demographic information, including the first 3 digits of his home zip code (ie, his city of residence), age, highest level of education attained, race/ethnicity, income, current work status, identity relating to sexuality, whether he had a domestic partner (if so, whether the partnership was registered), and his legal marital status. On completion, the interviewer recorded the interview end time.
All data are weighted. It was not possible to weight the data to adjust for demographic variables, such as age or education, because such statistics were not available for the total patron population. That probability weight is the multiplicative product of 3 weighting components: (1) the probability of a visit (the probability of a 3-hour shift being chosen for data collection multiplied by the probability of selection within the 3-hour shift), (2) the probability that a given respondent would visit the club during the data collection period (based on responses to survey questions on number of previous and subsequent planned visits to the venue during the 5-week data collection period), and (3) nonresponse/noncoverage (based on actual numbers of visits in a shift not matching the targeted number of visits).
The focus of the analytic approach was to identify significant correlates of engagement in risk behavior during the bathhouse visit, which we operationalized as anal intercourse without condoms. Proposed correlates were divided into 3 groups: (1) visit behavior (length of visit, had oral sex, had group sex, and felt effects of alcohol or other drugs), (2) prior behavior (number of men had sex with, sexual behavior in private spaces [private home or hotel], and sexual behavior in public spaces [including sex clubs/bathhouses] in the past 3 months; number of previous visits to target bathhouse and number of visits to other sex clubs in the past 12 months; current HIV serostatus; and tested for an STI in past 12 months), and (3) respondent characteristics (age, education, race/ethnicity, employment, income, self-defined orientation, marital status, domestic partnership, and zip code of residence). Bivariate relations between proposed correlates and outcome were evaluated using the SVY TABULATE procedure in STATA (Stata Corporation, College Station, TX),23 which calculates an F test based on the Rao and Scott second-order correction of the Pearson χ2 statistic.
Multivariate relations were assessed via logistic regression analysis using SPSS (SPSS, Chicago, IL) for model construction and the SVY LOGIT procedure in STATA23 to assess the final model. The proposed correlates were entered into the regression model hierarchically, with the visit behavior block entered first, the prior behavior block entered second, and respondent characteristics entered last. Using the logistic regression algorithm in SPSS, all correlates were entered stepwise within blocks, once using a forward stepwise procedure within each block and once using a backward stepwise procedure within each block. The analysis went through several iterations using successively more stringent statistical criteria for inclusion in the model. Ultimately, only correlates that achieved statistical significance (P < 0.10 on the likelihood ratio test for removal from the model) at the final step of the stepwise process (ie, with all eligible correlates entered in the model) were retained in the final model. The final model was then reanalyzed using the SVY LOGIT procedure in STATA to obtain corrected standard errors, P values for Wald statistics, and 95% confidence intervals (CIs) for odds ratios. Two models were assessed, 1 for the total sample (N = 440) and 1 for the subsample of men who reported being sexually active during the visit (n = 403). Both models achieved good fit, which was operationalized as P > 0.20 on the Hosmer-Lemeshow goodness-of-fit test.
Characteristics of Patrons
The sampling procedure selected 768 patrons as they exited the bathhouse. Of these, 40 were ineligible (ie, 25 were previously interviewed, 12 had insufficient English, and 3 were cognitively impaired) and 20 could not be interviewed (19 because an interviewer was not available and 1 because he exited the club too swiftly to be approached). Of the remaining 708 eligible patrons, 440 (62.1%) agreed to participate in the survey. Patrons who declined participation were not significantly different by age (P = 0.38) or race/ethnicity (P = 0.73) from those who agreed to participate (based on interviewer assessments of age and race/ethnicity).22 Table 1 shows the demographic characteristics, prior behavior, and visit behavior of the study participants and the bivariate relations between those correlates and unprotected anal intercourse (UAI) during the club visit.
Behavior During the Current Visit
The average visit at the bathhouse lasted 3.5 (median = 2.7, mode = 1.4, range: 0.3-17.5) hours. Overall, 55.8% of participants reported no anal sex during the visit, 33.0% reported protected anal sex only, and 11.1% (95% CI: 8.2% to 15.0%) reported UAI. During their visit, approximately 1 of every 12 respondents (8.4%) reported that he did not have any sex (ie, no anal or oral sex), whereas 91.5% reported that they engaged in oral sex during their visit. Table 2 shows insertive and receptive anal sex at the club, with and without a condom (ie, protected and unprotected, respectively). More men reported having insertive anal sex (28.8% protected only and 7.0% unprotected at least once) rather than receptive anal sex (13.9% protected only and 5.5% unprotected at least once). Among men who engage in insertive anal sex (n = 158), 19.6% (95% CI: 13.2% to 28.0%) reported doing so without a condom at least once. Among men who engaged in receptive anal sex (n = 85), 28.2% (95% CI: 18.7% to 40.1%) did so without a condom at least once. Among the 24 men who had receptive UAI, 13 (52.0%) were HIV-positive, whereas of the 31 men who had insertive UAI, 9 (29.0%) were HIV-positive.
Of the respondents who reported some oral or anal sex while at the bathhouse, the mean number of partners was 3.2 (median = 3.0, mode = 1.0, range: 1-30). The mean number of anal sex partners (among those who had anal sex) was 1.9 (median = 2.0, mode = 1.0, range: 1-20). Among those who had performed oral sex on other men, the mean number of other men was 2.5 (median = 2.0, mode = 1.0, range: 1-30). Among those who had partners perform oral sex on them, the average number of partners was 2.5 (median = 2.0, mode = 1.0, range: 1-20). Approximately 4 in 10 respondents (39.6%) engaged in group sex a mean number of 1.6 times during the visit (median = 1.0, mode = 1.0, range: 1-7) in group sizes (counting the respondent) averaging 3.7 men (median = 3.0, mode = 3.0, range: 3-20).
Approximately half of the men (51.7%) reported that they did not use any drugs (including alcohol) in the 24 hours before their visit. Although alcohol and marijuana were the 2 drugs most likely to be used in the previous 24 hours (31.3% and 12.0%, respectively), those men were not as likely to be feeling the effects of these 2 substances during their visit (23.2% and 52.9% of users, respectively). Conversely, all the other drugs mentioned (eg, amphetamines, poppers, GHB, Ecstasy, ketamine, Viagra) were almost universally felt while at the club (92.6%-100% of users). Of note, 17.3% of the respondents were responsible for the use of all these other drugs combined.
Behavior During the 3 Months Before the Current Visit
More than a quarter of respondents (27.5%) had UAI in the 3 months before their current bathhouse visit in a public space (ie, at a club or cruising area [6.0%]), private space (in a home or hotel room [15%]), or both [6.5%]). Many more men reported UAI in private (21.4%; 95% CI: 17.2% to 26.4%) than in public (12.5%; 95% CI: 9.5% to 16.3%) in the prior 3 months (P < 0.001), however. Approximately a quarter (24.3%) of the men who engaged in UAI in the previous 3 months also engaged in UAI during their visit to the club, whereas 5.4% of the men who did not engage in UAI in the previous 3 months engaged in UAI during their visit to the club (P < 0.0001).
Predictors of Sexual Risk in the Current Visit
We used multiple logistic regression analyses in 2 ways (Table 3). In the first analysis, we examined visit behavior, prior behavior, and respondent characteristics as predictors of engagement in UAI during the current visit among all men who went to the club (N = 440). In a second analysis, we narrowed the focus from all men in the club to only those men who reported sexual behavior (oral or anal sex) in the current visit (n = 403) to evaluate whether the number of sex partners during the current bathhouse visit might also predict UAI. The number of sex partners during the current bathhouse visit could not be included in the initial model because it was confounded with the outcome (ie, men with no partners cannot have UAI).
Among all men who went to the club, those who reported that they currently felt the effects of drugs used in the past 24 hours versus those who did not were more likely to report UAI during the current visit. Also, men who reported UAI in private or in public in the past 3 months and men who went to other clubs in the past 12 months were more likely to report UAI during the current visit. Additionally, men who were HIV-positive and who reported a domestic partner or spouse were more likely than their counterparts to report UAI during the current visit.
Focusing on only those men who had sex in the current visit, essentially the same significant predictors found in the initial analysis still obtain; however, several additional effects are also observed. Although a domestic partnership was no longer significant when the analysis was restricted to men who were sexually active during the current bathhouse visit, additional predictors of UAI during the club visit included the number of men the participant had sex with in the club and in the prior 3 months, accessing STI testing in the past 12 months, and income. Men who reported that they had sexual contact with 3 or more men during the current visit were more likely to engage in UAI during their current visit (18.2%) than men who reported sexual contact with 1 or 2 men (5.9%). The effect of the number of partners in the last 3 months was markedly different, however. We found that men who reported 3 to 5 partners during the prior 3 months had the highest likelihood of reporting UAI during the club visit. The difference in odds is significant when compared with men reporting 6 or more partners in the past 3 months but not significant when compared with men reporting fewer than 3 partners. Men who had an STI test in the past year and those with an income more than $100,000 compared with those with an income less than $40,000 were more likely to engage in UAI during the club visit.
Previous studies using ethnographies and convenience samples have presented a conflicting picture of sexual behavior that occurs in bathhouses,5-13 with some suggesting rampant risk behavior and others suggesting more limited risk. The use of probability sampling methodology in conducting the exit survey for the present study achieves a more precise estimate and clearer picture of sexual behavior in the setting and provides a standard against which others can be compared.
Most men at the bathhouse did not engage in risk behavior anywhere at all-not at the bathhouse, not at other sex settings, and not in private settings such as a home or hotel. Specifically, most men reported few partners during their visit, even with an inclusive definition of sex. Oral sex was the predominant sexual activity. Nearly everyone who had anal sex used condoms. Group sex was a marginal behavior, and when it did occur, it tended to take place in groups of 3. Surprisingly, a number of men who reported recent high-risk behavior in other settings actually reported safer sex during their current bathhouse visit. These study results contradict the assertion that the bathhouse facilitates large-scale high-risk sex.1,5
For those men who did engage in UAI, more reported insertive rather than receptive anal sex. This difference could be attributed to the social stigma of the “bottom” position,24 although it also could represent an attempt among these men to reduce their risk of infection, believing that insertive anal sex is less risky than receptive anal sex.25,26 Moreover, having measured anal sex with an inclusive definition resulted in men reporting anal sex even if the act was brief and did not lead to ejaculation; this behavior, referred to as “dipping,” is another way in which men might attempt to reduce their risk of HIV transmission.27-29
Because disease transmission does not require orgasm or even anal penetration, the concept of concurrency30 is particularly relevant in these settings, given that having multiple sex partners in rapidly formed sexual networks facilitates disease transmission, especially during disease outbreaks. For example, a single man having receptive oral sexual contact with 2 or more men during a visit, any of whom then have contact with even just 1 other partner, can serve as a conduit of disease transmission between men who did not have sex with each other. It is worth noting, however, that the data from this study suggest the extent of the networks in this setting are more limited than descriptions of these venues usually evoke;1,5 therefore, the transmission of disease is likely to be limited even in these settings.
Drug use frequently is seen as going hand-in-hand with bathhouse attendance,10,31 and men who use drugs are certainly more likely to go to bathhouses than men who do not use drugs.32 Men who take drugs and engage in sexual risk behavior represented a small proportion of the bathhouse customers, however; most men did not report feeling the effects of drugs during their bathhouse visit. Feeling the effects of drugs while at the bathhouse was significantly associated with sexual risk behavior, although the prevalence of men feeling the effects of a drug during their current visit was less than 1 in 5.
In this study, reported sexual risk behavior outside the bathhouse was instructive. Although it has long been known that men engage in more risk behavior in private settings than in bathhouses and other public sex venues,33 this difference usually has been dismissed because it is considered a consequence of men engaging in UAI at home with partners they know (assuming seroconcordance), whereas bathhouse partners are anonymous and cannot presume concordance. The present study found that UAI in private was not only more likely than UAI in public but that it was also predictive of UAI during the bathhouse visit. One explanation is that some men who engaged in UAI in private did so with men who were not necessarily better known to them (ie, men they met on the Internet, at private sex parties, at bars, in the street, or in any number of public settings). It may be that having sex at home or someone else's home setting allowed them to “feel” closer and to think they were more knowledgeable about their partners. Another possible explanation may simply be that once a man makes a decision to engage in UAI, he acts on that decision regardless of the environment. In any case, the fact that UAI in private and public settings was predictive of UAI at the bathhouse corroborates previous study results that show that eliminating bathhouses is not likely to eliminate the behavior,3,20,33 not because the behavior moves somewhere else but because that behavior already occurs somewhere else.
Bathhouses may vary from any given venue to the next; thus, generalizing from the study bathhouse to similar settings should be approached cautiously. The effects of different kinds of bathhouse environments on risk behavior are unknown but may differentially affect patron risk behavior or tend to draw a safer or riskier group of men.34 The participating club has many features and amenities that are typical in bathhouses throughout the United States and Canada, however.35 This was a cross-sectional study, and although data were collected at a single point in time, they are useful as a baseline measure against which other assessments at other times and at other clubs can be compared.
Self-reported data provide an opportunity for response bias. The men interviewed in this study could have downplayed or overstated their actual behavior. In a 1995 survey and observation study conducted at several sex clubs in which all sexual behavior was simultaneously observed,13 however, the observed and self-reported data were concordant, suggesting that men easily remembered and accurately reported their behavior in the venues. The most recent 3 months is a fairly short and memorable period and should result in accurate self-reported behavior. In addition, questions were broad (“any” experience rather than specific number of times engaging in specific behavior), as recommended for self-report about sexual behavior.36-38
A final limitation of note was that 40% of men refused to participate in the survey. It is not possible to know the degree to which the estimates would be altered if they had participated; however, it is likely that the variety of reasons men had for not participating had more to do with outside factors than with their experience and behavior at the club. For example, recruitment between 5:00 and 7:00 PM on weeknights was difficult, and many men reported that they could not participate because someone was waiting for them. Similarly, regardless of time, men who were leaving with a ride were less likely to agree to participate. An important reason why some men did not participate was their reluctance to be open about the fact that they go to a bathhouse; however, this group of men tends not to have risky sex and, correspondingly, has much lower rates of seropositivity.
The vast majority of the sexual activity that occurred in the setting would not contribute to HIV transmission, even though some men at the bathhouse persist in risk behavior not only at the bathhouse but in private and other public settings. The fact that men who engaged in UAI tended to do so wherever they had sex suggests that the bathhouse is a point of access to reach these men with appropriate intervention. The implication for prevention suggests, once again, that risk behavior is a complicated phenomenon. Reducing transmission among MSM is not about controlling individual behavior at the bathhouse or eliminating the venue altogether; many men who otherwise engage in high-risk behavior already report safer sex at the bathhouse. Instead, successful public health measures need to take advantage of the opportunity to promote direct, concrete, frank connections with the highest risk men in the population when they come through the bathhouse. More importantly, public health has the opportunity to design innovative prevention interventions that men can take with them and that target reducing their risk behavior not only at the bathhouse but wherever they have sex.
1. Shilts R. And the Band Played On: Politics, People and the AIDS Epidemic
. New York: St. Martin's Press; 1987.
2. Turner CF, Miller HG, Moses LE. AIDS: Sexual Behavior and Intravenous Drug Use
. Washington, DC: National Academy Press; 1989.
3. Binson D, Woods WJ, Pollack L, et al. Differential HIV risk in bathhouses and public cruising areas. Am J Public Health
4. Frankis J, Flowers P. Men who have sex with men (MSM) in public sex environments (PSES): a systematic review of quantitative literature. AIDS Care
6. Helquist M, Osmon R. Sex and the baths: a not-so-secret report. J Homosex
7. Rumaker M. A Day and a Night at the Baths
. San Francisco, CA: Grey Fox Press; 1979.
8. Tewksbury R. Bathhouse intercourse: structural and behavioral aspects of an erotic oasis. Deviant Behav
9. Elwood WN, Williams ML. Sex, drugs, and situation: attitudes, drug use, and sexual risk behaviors among men who frequent bathhouses. J Psychol Human Sex
10. Haubrich DJ, Myers T, Calzavara L, et al. Gay and bisexual men's experiences of bathhouse culture and sex: ‘looking for love in all the wrong places.’ Cult Health Sex
11. Weinberg MS, Williams CJ. Gay baths and the social organization of impersonal sex. Soc Probl
12. Richwald GA, Morisky DE, Kyle GR, et al. Sexual activities in bathhouses in Los Angeles County: implications for AIDS prevention education. J Sex Res
13. Woods WJ, Mayne T, Kegeles S. Few men at sex clubs and cruising areas have unprotected anal intercourse [poster]. Presented at: XI International Conference on Acquired Immune Deficiency Syndrome; 1996; Vancouver.
14. Van Beneden CA, O'Brien K, Modesitt S, et al. Sexual behaviors in an urban bathhouse 15 years into the HIV epidemic. J Acquir Immune Defic Syndr
15. McKusick L, Horstman W, Coates TJ. AIDS and sexual behavior reported by gay men in San Francisco. Am J Public Health
16. McKusick L, Wiley JA, Coates TJ, et al. Reported changes in the sexual behavior of men at risk for AIDS, San Francisco, 1982-1984-the AIDS Behavioral Research Project. Public Health Rep
17. Martin JL. AIDS risk reduction recommendations and sexual behavior patterns among gay men: a multifactorial categorical approach to assessing change. Health Educ Q
18. Martin JL. The impact of AIDS on gay male sexual behavior patterns in New York City. Am J Public Health
19. Bolton R, Vincke J, Mak R. Gay saunas: venues of HIV transmission or AIDS prevention? Natl AIDS Bull
20. Woods WJ, Binson D. Public health policy and gay bathhouses. J Homosex
21. Woods WJ, Binson D, Pollack LM, et al. Public policy regulating private and public space in gay bathhouses. J Acquir Immune Defic Syndr
22. Binson D, Blair J, Huebner D, et al. Sampling in surveys of lesbian, gay, and bisexual people. In: Meyer I, Northridge M, eds. The Health of Sexual Minorities: Public Health Perspectives on Lesbian, Gay, Bisexual and Transgender Populations
. New York: Springer; 2007.
23. Stata Corporation. STATA Survey Data Reference Manual Release 9
. College Station, TX: Stata Corporation; 2005.
24. Van de Ven P, Kippax S, Crawford J, et al. In a minority of gay men, sexual risk practice indicates strategic positioning for perceived risk reduction rather than unbridled sex. AIDS Care
25. Suarez T, Miller J. Negotiating risks in context: a perspective on unprotected anal intercourse and barebacking among men who have sex with men-where do we go from here? Arch Sex Behav
26. Vittinghoff E, Douglas J, Judson F, et al. Per-contact risk of human immunodeficiency virus transmission between male sexual partners. Am J Epidemiol
27. Wolitski RJ, Parsons JT, Gomez CA. Prevention with HIV-seropositive men who have sex with men: lessons from the Seropositive Urban Men's Study (SUMS) and the Seropositive Urban Men's Intervention Trial (SUMIT). J Acquir Immune Defic Syndr
. 2004;37(Suppl 2):S101-S109.
28. Hoff CC, Faigeles B, Wolitski RJ, et al. Sexual risk of HIV transmission is missed by traditional methods of data collection. AIDS
29. Orellana ER, Picciano JF, Roffman RA, et al. Correlates of nonparticipation in an AIDS prevention program for MSM. AIDS Educ Prev
30. Gorbach P, Holmes K. Transmission of STDs/HIV at the partnership level: beyond individual-level analyses. J Urban Health
. 2003;80(4 Suppl 3):iii15-iii25.
31. Elwood WN, Greene K, Carter KK. Gentlemen don't speak: communication norms and condom use in bathhouses. J Appl Commun Res
32. Stall R, Paul JP, Greenwood G, et al. Alcohol use, drug use and alcohol-related problems among men who have sex with men: the Urban Men's Health Study. Addiction
33. Bolton R, Vincke J, Mak R. Gay saunas: venues of HIV transmission or AIDS prevention? Natl AIDS Bull
34. Binson D, Woods WJ. A theoretical approach to bathhouse environments. J Homosex
35. Woods WJ, Binson D, Mayne TJ, et al. Facilities and HIV prevention in bathhouse and sex club environments. J Sex Res
36. Catania JA, Gibson DR, Chitwood DD, et al. Methodological problems in AIDS behavioral research: influences on measurement error and participation bias in studies of sexual behavior. Psychol Bull
37. Kauth MR, St. Lawrence JS, Kelly JA. Reliability of retrospective assessments of sexual HIV risk behavior: a comparison of biweekly, three-month, and twelve-month self reports. AIDS Educ Prev
38. Jaccard J, McDonald R, Wan CK, et al. Recalling sexual partners: the accuracy of self-reports. J Health Psychol
© 2007 Lippincott Williams & Wilkins, Inc.