Wong, Mee-Lian MPH, MD*; Chan, Roy K. W. MBBS, FRCP†; Koh, David MBBS, MSc(OM), PhD*
THE HIV EPIDEMIC IS GROWING rapidly in Asia.1,2 In 2005, an estimated 40.3 million people in the world are living with HIV with the steepest rise occurring in central and east Asia. The main mode of HIV transmission in Asia is heterosexual intercourse, largely related to the common practice of male patronage of female commercial sex workers.2 With increasing globalization, more men are traveling abroad to Asia and other countries, and a significant proportion among them engaged in casual sex (defined as sex with a previously unknown partner). Studies3–7 have investigated factors associated with their practice of casual sex. These included young age, travel without a spouse or partner, longer duration of stay, having multiple partners, use of illicit drugs, alcohol abuse, low perceived risk, and limited sexually transmitted infection (STI) knowledge. A significant proportion of travelers also engaged in unprotected sex with commercial sex workers.7–10 However, little is known about the reasons or factors associated with noncondom use with sex workers among these travelers. A study on tourists from the United Kingdom investigated the role of their personal characteristics on noncondom use with causal sexual partners in international beach resorts in Ibiza.5 Unprotected casual sex was found to be associated with smoking and higher numbers of sexual partners before their visit to Ibiza. The characteristics and behavioral factors of the sex workers may affect condom use with male travelers, but the possible influence of these factors has not been studied. Studies have been conducted on determinants of condom use among clients patronizing sex workers in their own country.11–16 They found condom use to be associated with characteristics of the clients rather than that of the sex workers. In general, clients who were higher educated11–13,15 paid higher charges,12 and had higher self-perceived risk13 were more likely to use condoms. It is not known, however, whether these same factors influence condom use among travelers patronizing sex workers in another country, where the sociocultural environment and pattern of sex work are often quite different.
Travelers are known to facilitate the spread of STIs17–19 and AIDS.20 Yet, there are few prevention programs for travelers, particularly those traveling from Asia. Health education programs have been set up for travelers from the West to promote safer sexual practices among them, but mixed outcomes have been reported from the few programs that have been evaluated. In a study by Gehring et al,21 safer sex information provided at the airport was not effective in modifying the sexual behavior of Swiss travelers. Recent research found that male travelers from Hong Kong were less likely to use condoms with sex workers in China15,16 than in their own country. In addition, HIV-related knowledge was not found to be predictive of condom use.16 These findings are disturbing because they suggest that general education campaigns may be ineffective in increasing condom use among male travelers who engaged in commercial sex abroad. The underlying reasons for low condom use among travelers, when they are abroad, need to be identified to help plan more effective programs to prevent them from getting HIV or STIs. If clients' knowledge is not associated with condom use, could the cultural milieu or the behavior of the sex workers in their country of travel influence condom use? There is paucity of studies on the influence of the sociocultural environment on condom use with sex workers among travelers when they go abroad and patronize sex workers. We therefore conducted a study to assess and compare sociodemographic, behavioral, and environmental factors associated with condom use in the most recent sexual act with a sex worker in and outside Singapore over the last 6 months among Singaporean men attending an STI clinic in Singapore.
Singapore is a small island republic with a population of approximately 3.5 million Singaporean residents in 2004. The majority of the residents are Chinese (76.0%) with 13.8% Malays and 8.4% Indians.22 It is situated in Southeast Asia with Indonesia and Malaysia being its neighboring countries. STI services in Singapore are run as a vertical program and are not integrated into the primary healthcare clinics. The Department of STI Control (DSC) clinic is the only public STI clinic in Singapore and it serves as the referral center for STI cases from other hospitals and private and government clinics. All male patients who had visited sex workers in the last 6 months either only abroad or only in Singapore and were attending (DSC) clinic from December 2001 to June 2003 for the first time for screening or treatment of STIs were recruited for the study. Clients who visited sex workers both abroad and locally in the last 6 months (n = 24, 5.7% of the total sample) were excluded for the purpose of this study.
Face-to-face interviews, using a structured questionnaire translated into Chinese and Malay, were conducted by 4 trained clinic staff in the privacy of the consultation rooms. The purpose of the study was explained to the respondent and consent obtained. The study was approved and given ethical clearance by the Department of STD Control and Ministry of Health. To reduce self-reporting bias, the interviewers stressed the confidentiality of the information and explained that the findings would be used to plan a better program to help protect them from AIDS and STIs. A free phone card worth U.S. $5 was given to each respondent at the end of the interview.
The primary outcome measure was condom use by the client in the most recent sexual contact with a sex worker in the last 6 months. The last sexual act was used in the analysis because this would ensure more accurate recall of the details of the sexual contact. The median time of the last sexual act from the day of the interview was 24 days (range: 1–180 days) for clients who visited sex workers in Singapore and 21 days (range: 2–180 days) for those who visited sex workers abroad.
With regard to the independent variables, we applied a comprehensive framework to assess clients' sociodemographic factors and behavioral and situational (environmental) factors associated with condom use in the last sexual act. The protection motivation theory,23 often used to analyze what motivates people to engage in behaviors that are usually unpleasant but would protect them from disease, was used to assess the behavioral factors in this study because condom use is often considered unpleasant by clients. Briefly, the protection motivation theory explains that a person is more likely to use condoms if he 1) perceives STI/AIDS to be serious, 2) feels his chance of getting the disease is high, 3) has high self-efficacy (confident in his ability to use condoms), and 4) believes that condoms are effective in protecting him from disease. We expanded the framework to include perceived barriers (potentially negative aspects) of using condoms such as finding it bothersome or nonpleasurable because these were commonly reported reasons for not using condoms. The previously mentioned behavioral factors and barriers were expressed as statements such as “Condoms will protect me from STIs and AIDS” and “I find it bothersome to use condoms.” Clients were asked to respond to the statements according to a 5-point Likert scale: strongly agree, somewhat agree, neither agree nor disagree, somewhat disagree, and strongly disagree. In addition, because we were studying clients who visited sex workers in other countries, environmental factors such as the local context of sex work with regard to 1) type of sex worker patronized such as whether she is operating from the brothels, streets, hotels, massage parlors, or other places of entertainment; and 2) whether she initiated condom use were also incorporated into the framework.
To ensure the cultural relevance of the questions, the first author also conducted in-depth interviews with 20 clients before the main study to explore their reasons for not using condoms, which were subsequently incorporated into the questionnaire. Respondents were asked to give a “yes” or “no” response to each reason.
All analyses were done separately for clients patronizing sex workers in and outside Singapore. In the univariate analysis, χ2 tests were used to compare proportions in condom use for categorical variables and independent sample t test to compare means for continuous variables. To protect against an inflated α and hence a type 1 error, when we conclude that there is a significant difference when there is none, we used a Bonferroni correction. We divided our level of significance set at 0.05 by the number of comparisons in the univariate analysis. Because 20 comparison tests were done, a comparison was only considered statistically significant if it has a significance level of 0.0025.
To identify factors significantly associated with condom use (yes or no) after controlling for potential confounders, multivariable analysis, using a modification of Cox proportional hazards regression model24,25 for cross-sectional data, was used. All independent variables with a statistical significance of 0.1 or less with condom use in univariate analysis were entered into the regression model, using backward stepwise elimination procedure, with condom use (yes or no) as the dependent variable. This model yields the adjusted prevalence ratios by independent variables simultaneously adjusted for all other confounding variables. Although the multiple variable logistic regression model is often used to control for confounding in the analysis of the association of multiple independent variables with a single dependent dichotomous variable, this model has a drawback in that it yields the odds ratio as a measure of association. The odds ratio is less interpretable, and it only approximates the prevalence ratio well provided the dependent variable is a rare event. Because condom use in our study is not a rare event, we needed a model that can directly estimate the prevalence ratio, yet allow for adjustment of confounding. Breslow26 has shown that by assuming constant risk period, namely the “persons at risk cohort,” the conditional hazard ratio estimated by Cox regression model can be adapted to estimate prevalence ratios for cross-sectional data. The adjusted prevalence ratios of condom use and the 95% confidence intervals were calculated based on the estimated regression coefficients from the regression model. Statistical significance in the cox regression analysis is set at P <0.05. All data analyses were performed with the Statistical Package for Social Science (SPSS), version 13.0 (SPSS, Chicago, IL).
Sociodemographic Characteristics of the Study Sample
Three hundred ninety-six (99.0%) of the 400 men recruited for the study gave informed consent. Twenty-four (6.0%) men were excluded from analysis because they could not recall whether they used condoms during their last sexual act. They tended to be single (79.2% vs. 59.9%, P <0.05) younger (mean age: 30.0 vs. 35.0 years, P <0.05), and were more educated (87.5% vs. 76.9%, P <0.05). A total of 372 clients were analyzed.
Comparison Between Men Who Patronized Sex Workers In and Outside Singapore
Almost equal proportions (51.6% and 48.4%), respectively, visited sex workers in Singapore and abroad. Countries most often visited by the clients were Indonesia (57.2%), Thailand (16.7%), Malaysia (14.4%), and China (2.8%). Only 1.7% of the clients visited sex workers in the West. The 2 groups did not differ significantly in marital status, educational level, race, occupation, or income (Table 1). Clients in Singapore were, however, younger and the majority patronized sex workers from brothels. In contrast, quite similar proportions of clients abroad visited sex workers from brothels and massage parlors, and more of them compared with clients in Singapore patronized sex workers from places of entertainment.
Comparison of Factors Associated With Condom Use In and Outside Singapore
Almost all the clients (87.5%) in Singapore used condoms in their last sexual contact with the sex worker compared with only 54.4% (P <0.001) of the clients who patronized sex workers overseas. Condom use in other countries varied from 40.8% to 77.0% with the lowest percentage of condom use being reported by clients (n = 103) who visited China (60.0%) and Indonesia (40.8%).
Table 2 compares the univariate analysis of factors associated with condom use between clients who patronized sex workers in and outside Singapore. Both groups generally showed similar relationships. Condom use among clients in and outside Singapore was significantly associated with sex workers' initiation of condom use. For clients in Singapore only, condom use was significantly associated with age, income, perceived risk, confidence in using condoms, and perceived barriers to condom use such as bother or reduction in sexual pleasure. For clients outside Singapore only, condom use was significantly associated with educational level and perception of whether their friends used condoms. For both groups, condom use was not significantly associated with marital status, race, occupation, alcohol drinking or smoking status, perceived seriousness of AIDS or STIs, perceived efficacy (effectiveness) of condoms in protecting against STIs and AIDS, and liking for or ease of condom use.
Multivariable Analysis of Condom Use
Table 3 shows the univariate and the adjusted prevalence ratios of condom use in the final models following stepwise Cox regression analysis. After controlling for confounding, the only factor significantly associated with condom use in the last sexual contact for both groups of clients was initiation of condom use by the sex worker. Clients were approximately twice as likely to use condoms if the sex workers initiated condom use. Those with higher education were also more likely to use condoms, but this was found only among clients who visited sex workers outside Singapore. All the other significant associated factors on univariate analysis were no longer associated with condom use after controlling for confounders in the multivariable analysis.
Reasons for Not Using Condoms
Reported reasons for not using condoms by clients visiting sex workers outside Singapore were collapsed into 2 groups: 1) extrinsic reasons related to the situational context of the sexual act such as nonavailability of condoms (23.1%), sex workers did not ask (20.7%), heat of the moment (19.5%), and drunkenness (19.5%); and 2) intrinsic reasons related to the clients such as reduced sexual pleasure (30.5%), trusted the sex worker (6.1%), could not do it with a condom (4.9%), did not use out of habit (3.7%), and did not know how to use (2.4%). Extrinsic reasons were reported by the majority (69.5%) of the clients. Less than half (43.9%) gave intrinsic reasons. Similar reasons were found for clients in Singapore.
Availability of Condoms In and Outside Singapore
Condoms were not available or provided to clients in a significantly higher proportion of sexual acts with sex workers abroad than locally (26.2% vs. 12.5%, P <0.001) (Fig. 1). In addition, a significantly lower proportion of sex workers abroad (32.7% vs. 75.6%, P <0.001) initiated condom use and provided condoms. Less than one fourth of clients in Singapore and abroad brought their own condoms.
The only independent factor significantly associated with condom use for clients both in and outside Singapore after multivariable analysis was initiation of condom use by the sex workers. Condom use was not significantly associated with the clients' age, martial status, race, occupation, income, knowledge of the seriousness of AIDS or STIs, effectiveness of condoms, self-perceived risk, self-efficacy in condom use, and their dislike for condoms.
To our knowledge, our study is the first to apply a behavioral and environmental framework to assess determinants of condom use between travelers and sex workers. Previous studies did not assess whether sex workers negotiated condom use with the clients.3,15,16 We found that condom use was determined more by extrinsic factors such as the sex workers' actions rather than the clients'. Regardless of the geographic location, clients were approximately twice more likely to use condoms if asked by the sex workers. Our study found that condom use was not associated with intrinsic personal factors of the clients such as their awareness of seriousness of AIDS, self-perceived risk, self-perceived effectiveness of condoms, self-efficacy in using condoms, and dislike for condoms. Reported reasons for not using condoms were also mainly extrinsic in nature because main reasons given were nonavailability of condoms, sex workers did not ask, done at the heat of the moment, and drunkenness. This lack of association of condom use with personal factors such as the clients' knowledge, self-perceived vulnerability, and attitudes on condom use might explain why client education16,21 was ineffective in increasing condom use among men who traveled abroad for commercial sex.
The low condom use among Singaporean clients with sex workers in other countries compared with sex workers in their own country could therefore be explained by differences in the local context of the sexual transaction. Condom use was initiated by only one third of sex workers in countries outside Singapore compared with the majority (76%) of the sex workers in Singapore. All brothel-based sex workers and masseuses in Singapore were trained to negotiate condom use with their clients, and owners of these establishments were required to support the sex workers in this. Details of the intervention program are described elsewhere.27,28 Briefly, the brothel-based sex workers were trained how to negotiate condom use through video presentations, role play, small group discussions, and educational pamphlets. Condom negotiation techniques used by the successful sex workers were used to design messages for the video presentations and educational pamphlets on condom negotiation techniques.29 Posters on 100% condom use were also distributed to all brothels, and the Department of STD Control checked on the brothels' compliance to ensure that the posters were displayed prominently in the waiting area, toilets, and bedrooms. Condoms were provided free to clients and displayed prominently at the reception counters. Brothels with high gonorrhea incidence rates of more than 10% were also temporarily suspended from business. In contrast, sex workers in many Asian countries do not have the support of their brothel managers or owners of sex establishments to use condoms.30 It is also likely that these sex workers do not have the skills to negotiate condom use.
The other factor significantly associated with condom use with clients visiting sex workers abroad was higher educational level of clients. Further analysis found that sex workers initiated condom use with significantly higher than lower-educated clients (37% vs. 20%, P <0.05). Initiation of condom use did not differ by the type of sex workers, that is, with regard to whether they were brothel-based, street based, or recruited from hotels or massage parlors. It is likely that sex workers patronized by higher-educated clients were higher class and educated with better access to condoms and hence are more likely to use condoms.
Could the geographic difference in condom use be explained by differences in the sociodemographic characteristics between the clients who engaged in commercial sex in Singapore compared with those who did so overseas? The 2 groups of clients did not differ in educational level. Although they differed significantly in age and type of sex workers patronized, these characteristics were not found to be significantly associated with condom use after controlling for confounders in the multivariable analysis. Hence, the geographic difference in condom use is unlikely to be the result of the differences in the characteristics of the 2 groups of clients.
Our study has some limitations. Patients attending an STI clinic may not be representative of the general traveling Singaporean public who visited sex workers abroad. However, the strength of our study was that we could conduct a comprehensive behavioral survey in a clinical setting to assess determinants of condom use. The most recent sexual act with the sex worker might not represent the usual practice for the client. However, it facilitated accurate recall and allowed us to study specific details of the most recent sexual transaction with the sex worker.
Public Health Implications of Our Study
Our results have important implications for HIV prevention among travelers. Because condom use by clients was prompted by sex workers and was not intrinsically motivated, they may adopt riskier practices of noncondom use in a different geographic location if sex workers did not initiate condom use. HIV prevention programs targeting travelers going abroad for commercial sex should go beyond information dissemination to increase their access to condoms and to get sex workers to remind or negotiate with them to use condoms at that crucial point of sexual contact. In addition to working with travel agents, we need to collaborate with the entertainment industry, hoteliers, massage parlors, and brothel managers through government or nongovernment organizations in other countries to increase the availability of condoms and to empower sex workers to negotiate condom use.
It may be argued that equipping sex workers with condom negotiation skills may not solve the problem because there have been reports of clients in Asia who were willing to pay more for sex without a condom or who threatened to use violence when sex workers insisted on condom use.2 The proportion of clients resorting to such actions was not reported. It may well represent a minority rather than the majority of clients. Our study found that the majority (69.5%) of the clients did not use condoms because of extrinsic reasons such as nonavailability, sex workers did not ask, drunkenness, or heat of the moment. These factors are more within the control of the sex workers than the clients. If sex workers could offer, persuade, and remind them to use condoms at that point of time or just put it on their drunk client in a pleasurable way, as was practiced by sex workers in Singapore, it would increase the chance of clients using condoms. There is supportive evidence that sex workers could persuade clients to use condoms. A noteworthy finding in this current study was that slightly more than half (52.8%) of the clients with dislike for condoms actually used condoms after persuasion by the sex workers. A recent study on brothel-based sex workers in Indonesia30 also found that approximately one third of clients (36.5%) who initially refused condoms used them after persuasion by the sex workers. Our experience in Singapore found that we could increase condom use among sex workers from less than 50% to more than 90% after teaching them negotiation skills and mobilizing brothel support.27,28 Although we noted the differences in the working environment of sex workers in Indonesia and Singapore, the previously mentioned findings gave us optimism that clients in Asia may be persuaded to use condoms if sex workers could be empowered and supported by brothel management to negotiate condom use with them at that crucial point of sexual contact. Point-of-decision prompts have been found to be effective in changing people's lifestyle behavior such as getting them to take the stairs to increase physical activity.31
In conclusion, lower condom use among Singaporean men who traveled to other Asian countries for commercial sex was associated with the local milieu, whereby sex workers in these countries were less likely to initiate condom use. Further research should be conducted to assess whether the same clients practiced differential condom use with sex workers in their own country and abroad. In addition, studies should be conducted on travelers from the West to Asia to determine the role of environmental and personal factors in influencing condom use between them and the sex workers.
1. Ruxrungtham K, Brown T, Phanuphak P. HIV/AIDS in Asia. Lancet 2004; 364:69–82.
2. Joint United Nations Program on HIV/AIDS (UNAIDS) and World Health Organization (WHO). AIDS Epidemic Update—December 2005: Geneva: UNAIDS/WHO, 2005.
3. Abdullah ASM, Hedley AJ, Fielding R. Travel, sexual behaviour and the risk of contracting sexually transmitted diseases. Hong Kong Med J 1998; 4:37–44.
4. Gillies P, R Slack, Stoddart N, Conway S. HIV-related risk behaviour in UK holiday makers. AIDS 1992; 6:339–341.
5. Bellis MA, Hughes K, Thomson R, et al. Sexual behaviour of young people in international tourist resorts. Sex Transm Infect 2004; 80:43–47.
6. Matteelli A, Carosi G. Sexually transmitted diseases in travellers. Clin Infect Dis 2001; 32:1063.
7. Abdullah ASM, Fielding R, Hedley AJ, et al. Risk factors for sexually transmitted diseases and casual sex among Chinese patients attending government sexually transmitted disease clinics in Hong Kong. Sex Transm Dis 2002; 29:360–365.
8. Hawkes S, Hart GJ, Johnson AM, et al. Risk behaviour and HIV prevalence in international travellers. AIDS 1994; 8:247–252.
9. Mulhall BP. Sex and travel: Studies of sexual behaviour, disease and health promotion in international travellers. Int J STD AIDS 1996; 7:455–465.
10. Kleiber D, Wilke M. AIDS related risk behaviour in sex tourists in Thailand. In: Heckmann W, Koch MA, eds. The AIDS Era. Berlin: Sigma Rainer Bohn Verlag, 1994.
11. Estebanez P, Rua-Figueroa M, Aguilar MD, et al. Condom use in clients of prostitutes in Spain. AIDS 1996; 10:445–446.
12. Pickering H, Quigley M, Hayes RJ, et al. Determinants of condom use in 24,000 prostitute/client contacts in the Gambia. AIDS 1993; 7:1093–1098.
13. Graaf RD, Gertjan VZ, Vanwesenbeeck I, Straver CJ, Visser JH. Condom use by Dutch men with commercial heterosexual contacts: Determinants and considerations. AIDS Educ Prev 1997; 9:411–423.
14. Wee S, Barrett M, Wong ML, et al. Determinants of inconsistent condom use with female sex workers among men attending the STD clinic in Singapore. Sex Transm Infect 2004; 80:310–314.
15. Lau JT, Amy Tang SY, Tsui HY. The relationship between condom use, sexually transmitted diseases and location of commercial sex transactions among male Hong Kong clients. AIDS 2003; 17:105–112.
16. Lau JT, Siah PC, Tsui HY. Behavioural surveillance and factors associated with condom use and STD incidences among the male commercial sex client population in Hong Kong, results of two surveys. AIDS Educ Prev 2002; 14:306–317.
17. Rogstad KE. Sex, sun, sea and STIs: Sexually transmitted infections acquired on holiday. BMJ 2004; 329:214–217.
18. Centers for Disease Control and Prevention. Increases in fluoroquinolone-resistant Neisseria gonorrhoeae
—Hawaii and California, 2001. JAMA 2002; 288:2961–2963.
19. Hawkes S, Hart GJ, Bletsoe E, Shergold C, et al. Risk behaviour and STD acquisition in genitourinary clinic attenders who have travelled. Genitourin Med 1995; 71:351–354.
20. Vittecoq D, Roue RT, Mayaud C, et al. AIDS after travelling in Africa: An epidemiological study in seventeen Caucasian patients. Lancet 1987; ii:612–614.
21. Gehring TM, Widmer KJ, Keliber D, Steffen R. Are preventive HIV interventions at airports effective? J Travel Med 1998; 5:205–209.
23. Rogers RW, Prentice-Dunn S. Protection Motivation Theory. In: Gochman DS, ed. Handbook of Health Behaviour Research 1: Personal and Social Determinants. New York: Plenum Press, 1997:113–132.
24. Axelson O, Fredrikson M, Ekberg K. Use of the prevalence ratio versus the prevalence odds ratio as a measure of risk in cross-sectional studies [Letter]. Occup Environ Med 1994; 51:574.
25. Lee J. Odds ratio or relative risk for cross-sectional data. Int J Epidemiol 1994; 23:201–203.
26. Breslow NE. Covariance analysis of censored survival data. Biometrics 1974; 30:89–99.
27. Wong ML, Chan RKW, Koh D. The long term effects of condom promotion programmes for vaginal and oral sex on sexually transmitted infections among sex workers in Singapore, 1990–2002. AIDS 2004; 18:1195–1199.
28. Wong ML, Chan R, Koh D, Wong CM. Theory and action for effective condom promotion. Illustrations from a behaviour intervention project for sex workers in Singapore. International Quarterly of Community Health Education 1995; 15:405–421.
29. Wong ML, Archibald C, Chan R, Goh A, Tan TC, Goh CL. Condom use negotiation among sex workers in Singapore: Findings from qualitative research. Health Education Research: Theory & Practice 1994; 9:57–67.
30. Basuki E, Wolffers I, Walter D, et al. Reasons for not using condoms among female sex workers in Indonesia. AIDS Educ Prev 2002; 14:102–116.
31. Task Force on Community Preventive Services (Task Force), Centers for Disease Control and Prevention (CDC). Guide to Community Preventive Services. Point-of-Decision-Prompts That Encourage People to Use the Stairs Are Recommended to Promote Physical Activity, 2004. Available at: www.thecommunityguide.org/pa/pa-int-decision-prompts
. Accessed February 24, 2005.