THERE IS EVIDENCE that orogenital sex transmits gonorrhea, syphilis, chlamydia, chancroid,1 and oral human papillomavirus infections.2 Protection for all episodes of intercourse, including oral sex, has been recommended by public health authorities since the early 1980s.
Oral sex, which was once considered an uncommon sexual behavior, has become a relatively common practice. A study in London on women attending genitourinary medicine clinics reported a marked increase in the prevalence of oral sex from 69.7% in 1982 to 81.9% in 1992 (for penetration) and from 36.9% in 1982 to 51.4% in 1992 for ejaculation.3 In Singapore, oral sex among female brothel-based sex workers has also increased sharply from 27.1% in 1990 and 1992 to 81.7% in 1997.4 This increase, which occurred across all ages and ethnic groups of the sex workers, could partly be caused by clients' demands for other alternatives, after the successful implementation of a condom promotion program for vaginal sex. The upward trend in oral sex has been accompanied by a significant concomitant increase in pharyngeal gonorrhea from 0.69% of brothel-based sex workers in 1992 to 1.31% in 1996.4
The use of condoms for oral sex varied considerably from one country to another with a high rate of 92% reported from local female sex workers in Sydney5 and low rates of < 20% among male sex workers6 and none among female sex workers in Indonesia.7 The extent of consistent condom use for oral sex among female brothel-based sex workers in Singapore is unknown. However, consistent condom use for vaginal sex among them is high (97%) as a result of a comprehensive condom promotion program for vaginal sex. After this project, consistent condom use for vaginal sex with paying clients increased markedly from 44.4% in 1994 to 73.6% at 1 year and 90.5% at 2 years with a concomitant decline in cervical gonorrhea from 2.85% in 1994 to 0.7% in 1996.8 In contrast, pharyngeal gonorrhea among brothel-based sex workers showed an upward trend that paralleled the marked increase in oral sex. The increase in pharyngeal gonorrhea in contrast to a sharp decline in cervical gonorrhea suggests that condom use for oral sex may not be high as for vaginal sex.
This study aims to determine the prevalence of and factors associated with consistent condom use for oral sex among female brothel-based sex workers with paying clients in Singapore. The findings would be used to plan an intervention to increase consistent condom use for oral sex and hence reduce oral sexually transmitted infections or prevent AIDS.
Background Information of Brothel-Based Sex Workers in Singapore
In Singapore, all brothel-based sex workers are registered with the Medical Surveillance Scheme8 at the Department of STD Control (DSC). The Scheme requires all brothel-based sex workers to maintain contact with the sexually transmitted disease (STD) clinic at this department and designate general practice clinics for their regular monthly screening for gonorrhea and chlamydia infections and three monthly screenings for HIV and syphilis. The Anti-Vice Unit works closely with the DSC to ensure that the sex workers turn up regularly for their screening tests. Sex workers found to have STDs are treated immediately to prevent dissemination of the disease. Since 1992, the DSC has routinely conducted talks and skills development classes to all brothel-based sex workers to promote consistent condom use. Condoms are easily available in Singapore and sex workers provide condoms to their clients. Regular talks are also given to brothel keepers to ensure that they support the sex workers in consistent condom use.
All brothel-based sex workers are registered with the Department of STD control for HIV/STD surveillance and each of them had a medical record of her sexual behavior and screening test results. A sample of 250 records was randomly selected from the register of those female brothel-based sex workers with a history of oral sex documented in their records (n = 800) within the last 1-year period. The sample size was calculated to give a 95% confidence level (CI) and 5% error tolerance to determine the prevalence of consistent condom use for oral sex, which was estimated to be 60%.
The term oral sex refers to receptive oral sex in which the client's penis is inserted into the sex worker's mouth. Trained interviewers went to the brothels to conduct the interviews from June to December 1998. Twenty-five (10%) sex workers claimed that they no longer practiced oral intercourse. They were excluded from the study. Finally, 225 respondents were interviewed with no refusals.
The questionnaire included questions on sociodemographic characteristics and factors influencing the behavior of consistent condom use. One component of Green's PRECEDE-PROCEED framework9 and our previous research on factors associated with condom use for vaginal sex10 were used to select factors influencing consistent condom use in our study. Greens PRECEDE-PROCEED framework is a health education/promotion planning framework that integrates insights from other models to guide the development of health promotion or health education intervention programs to affect behavioral and environmental change. One of the components of the model is the analysis of factors influencing behavior change. In this analysis, Green integrates some established theoretical or explanatory models of behavior such as Becker's Health Belief Model11 and Bandura's Social Learning Theory,12 which has a component on self-efficacy, and he subsequently classifies factors influencing behavior, for example, condom use into three categories, namely (1) predisposing factors such as knowledge, self-perception, and attitudes that motivate the person to use condoms; (2) enabling factors referring to skills, time, or environment that enable her to use condoms; and (3) reinforcing factors referring to encouragement and support from peers.
Before designing the questionnaire, the first author conducted in-depth interviews with 30 sex workers, using open-ended questions to understand better sex workers' perspectives on oral sex and condom use and to explore the range of reasons for using or not using condoms during oral sex. This information was used to design more appropriate questions on factors influencing condom use. Questions on pre-disposing factors included sex workers' knowledge on transmission of AIDS by oral sex, perception of their right to refuse sex without a condom, and their vulnerability of getting AIDS or STDs. Questions on enabling factors included the level of their technical or negotiation skills to get client to use condoms and their attendance at skill development classes where they were taught negotiation skills for condom use. Their negotiation skills were assessed by their self-reported success rate in getting clients to use condoms after persuasion and their technical skills were assessed by their ability to put the condom on the client's penis with their mouth. The latter skill was reported to the first author by sex workers during the in-depth interviews as an effective method of putting the condom on the client's penis without resistance as he enjoyed it. Questions on nonenabling factors or barriers to condom use, such as the unpleasant taste of condoms, were also asked. Reinforcing factors included encouragement by their peers to use condoms and self-perceptions on whether their friends used condoms. We also asked about the ethnic group of their clients because earlier in-depth interviews showed that sex workers would use condoms more often with clients such as Malays, Indonesians, Indians, and Westerners because they perceived these clients to be higher risk. These clients were also reported to be persuaded more easily.
Consistent condom users for oral sex refer to sex workers who used condoms during every act of oral sex with clients. These would include those sex workers who always refused oral sex with clients who could not be persuaded to use condoms.
The face-to-face interviews were conducted by the first author and trained interviewers. To reduce self-reporting bias, the interviewers stressed the confidentiality of the information and explained that the results would be used to plan a better program to help protect the sex workers from STD infections. They also explained to the sex workers on their awareness of their difficulties in getting clients to use condoms for oral sex and that “100% condom use” answers were not expected. Sex workers were encouraged to tell the truth. Questions were also translated into simple, colloquial terms with which the sex workers were familiar. The question on condom use was asked in an open, non-threatening and non-judgmental format as follows: On average, how many out of 10 clients do you use condoms for oral sex? This question has been pretested, used in our previous surveys, and found to be well understood by the sex workers. Earlier pretesting has also shown that that sex workers were less likely to overreport condom use with this question as compared to the question: Do you always use condoms for oral sex with your clients? Sex workers could also relate to units of 10 because they see an average of 10 clients per day. The clinical records of the sex workers were also checked for pharyngeal gonorrhea in the past year.
Chi-square test was used to compare the differences in proportion among groups. Chi-square trend was used for categorical variables ranked on an ordinal scale. Independent sample t-test was used to compare means for continuous variables such as age, mean number of clients, and mean self-reported success rate in getting clients to use condoms by condom use. Because this is a cross-sectional study, the association of consistent condom use (dependent variable) with the independent variables was evaluated by the prevalence rate ratio. Multivariate analysis, using a modification of Cox's proportional hazards regression model for cross-sectional data13,14 was used to determine factors that remained significantly associated with consistent condom use after controlling for potential confounders. Factors significantly associated with consistent condom use in univariate analysis (P < 0.05) were entered into the regression model with consistent condom use as the dependent variable. For all the variables, an estimated prevalence rate ratio of consistent condom use and a 95% CI were calculated, based on the estimated coefficients from the regression model.
Table 1 shows the sociodemographic characteristics of the sex workers. Most were single with a mean age of 28.5 years (range 22-56 years), with almost equal distribution by class. The most common clients seen were local Chinese (reported by 99.6% of sex workers), followed by Malays (43.1%), Japanese (39.1%), and Indians (21.3%). Less than 10% reported entertaining Bangladeshi, Indonesian, or Thai clients. Approximately half of the sex workers would only see Chinese, Japanese, or Taiwanese clients and would not entertain clients such as Malays, Indonesians, Indians, Bangladeshis, or whites because they were perceived to be high-risk clients.
Prevalence of Consistent Condom Use
Slightly more than half (56.9%) consistently used condoms during oral sex with clients (Table 1). In contrast, almost all (97%) used condoms consistently for vaginal sex. Only 1.5% practiced anal sex and none practiced oral anal contact. Most always asked their clients to use condoms during oral sex (82.7%) and vaginal sex (98.6%). Sixteen percent did not ask their clients to use condoms for oral sex, but used their mouth to put the condom on their client's penis. The success rate in getting clients to use condoms after negotiation was higher for vaginal sex than oral sex (mean 97.0% vs. 72.8%, median 100% vs. 80%, P < 0.01). Overall, a significantly higher percentage of sex workers knew that AIDS can be transmitted by unprotected vaginal sex than unprotected oral sex (89.1% vs. 74.4%, P < 0.001).
Factors Associated With Consistent Condom Use
Table 2 shows the percentage of consistent condom users by sociodemographic and behavioral characteristics. Consistent condom use was significantly associated with class and race of client. Sex workers who were middle class or had Malay, Indonesian, Bangladeshi, or Indian clients were significantly more likely to use condoms than those who were high class or saw exclusively Chinese, Japanese, or Taiwanese clients. Consistent condom users appeared to have more clients than inconsistent condom users, but the differences were not statistically significant (mean clients/day: 10.3 vs. 9.4, P > 0.05).
Consistent condom use was also analyzed by three categories of behavioral factors, namely predisposing, enabling, and reinforcing factors. It was not significantly associated with knowledge of transmission of AIDS by oral sex, awareness of their right to safe sex, their perceived vulnerability to getting AIDS by oral sex, and the taste or smell of condoms. However, it was significantly associated with self-efficacy or belief in their ability to get their clients to use condoms and enabling factors such as negotiation skills (success rate in getting clients to use condoms after persuasion), skills in putting the condom on the penis with their mouth, and attendance at skills development classes. A dose-response relationship was seen in which the percentage of consistent condom users increased with increased proportions of clients successfully persuaded to use condoms. Consistent condom use was also significantly associated with encouragement by peers to use condoms and self-perception of condom use among peers.
The combined effect of all the variables significantly associated with consistent condom use (P < 0.05) was assessed in Table 3. Class and negotiation skills of sex workers were found to be significantly associated with condom use. Middle class sex workers were 1.5 times more likely than high class sex workers to use condoms consistently. Sex workers who succeeded in persuading all their clients to use condoms were twice more likely to use condoms consistently compared with sex workers who could not get all their clients to use condoms. Race of client was also marginally significant with consistent condom use.
Further analysis was carried out on high and middle class sex workers, to identify factors that could explain the difference in condom use between them (Table 4). High class sex workers were younger and more educated. They did not differ from middle class sex workers in knowledge of transmission of AIDS by oral sex. However, they were significantly less likely than middle class sex workers to believe in their ability to get their clients to use condoms (self-efficacy). This was supported by their lower mean success rate in getting clients to use condoms. Only 18% of them could get all their clients to use condoms compared to 39% among middle class sex workers. High class sex workers were also less likely to be encouraged by their peers to use condoms.
Pharyngeal Gonorrhea and Condom Use
The records of the sex workers were checked for HIV and STD infections over the past year. None of the sex workers had HIV. The cumulative annual incidence of pharyngeal gonorrhea episodes per sex worker for the year 1997 was slightly higher than cervical gonorrhea, but the difference was not statistically significant (4.0% vs. 3.1%, P = 0.400). When pharyngeal gonorrhea was compared by consistent condom use, the cumulative annual incidence of gonorrhea episodes per sex worker was 4.62 times higher among inconsistent condom users than consistent condom users (7.2% vs. 1.6%, P < 0.05).
Slightly more than half (56.9%) of brothel-based sex workers in Singapore used condoms consistently for oral sex, and this is much lower than consistent condom use for vaginal sex. Our findings concur with a study on female crack-smoking sex workers in the United States that found lower consistent condom use for oral sex (31%) than vaginal sex (48%).15 However, our findings differ from Alary's study in Copenhagen, which found consistent condom use to be equally high at almost 100% for vaginal and oral sex.16
The lack of correlation of perceived vulnerability with consistent condom use concurs with the findings on female sex workers in the Philippines17 and could be explained by the cross-sectional design of this study, which does not allow us to assess the direction of the causal effect relationship. It is unknown whether the self-perception of vulnerability precedes or is a consequence of the behavior. Condom-using sex workers might perceive themselves to have low risk of getting AIDS rather than the perception of high risk leading to condom use. Self-perceived vulnerability is also influenced by psychological and cognitive factors. Sex workers might rationalize their risk as low to cope with the anxiety of not being able to do anything about condom use. Knowledge on the transmission of AIDS by unprotected oral sex was also not associated with increased condom use. The knowledge practice gap could be explained by other factors coming between knowledge and practice such as lack of skills to negotiate condom use for oral sex and fear of losing clients.
Negotiation skill was the most important factor significantly associated with condom use in our study. Sex workers who could persuade all their clients to use condoms were about twice more likely to use condoms consistently compared to their peers who did not succeed in persuading all their clients to use condoms. Another notable finding is that high-class sex workers, although better educated, were less successful than middle class sex workers in negotiating condom use. The clients of the former were probably more demanding and less easily persuaded because they paid more for the services of the sex workers.
Our study has some limitations. Sex workers were selected from a register of sex workers whose medical records indicated history of oral sex. This sampling frame might have excluded those sex workers who denied oral sex. However, the selection bias is unlikely to be gross because the history was taken by trained, specialized, and experienced health staff at the DSC. These staff who also counsel sex workers and patients with HIV have been trained on history taking of sexual behavior. In addition, 10% of the sex workers in our sample subsequently reported that they no longer practiced oral sex and were left out of the survey. The non-participation bias would have led to an overestimation of condom use if the non-participants were not condom users. The cross-sectional study design also makes it difficult to tell the direction of the relationship between some variables such as vulnerability and condom use. Sex workers may overreport consistent condom use for oral sex, but we have taken steps to minimize this bias. We did not consider other factors influencing consistent condom use for oral sex, such as clients' characteristics. Finally, our study findings cannot be extrapolated to street-based sex workers for whom working conditions might differ. However, the strength of the study is that we look at the comprehensive set of factors influencing condom use, which were derived from established explanatory models of behavior and from in-depth interviews of sex workers' perspectives on condom use.
In conclusion, consistent condom use for oral sex among female brothel-based sex workers was much lower than for vaginal sex, and it was significantly associated with class of sex workers and negotiation skills, but not with their knowledge of transmission of AIDS by unprotected oral sex. Future interventions should provide training to sex workers, particularly high class ones, on negotiation skills for oral sex. In addition, the finding on higher pharyngeal gonorrhea rates among inconsistent condom users should be fed back to health staff and sex workers to encourage condom use during oral sex.
1. Edwards S, Crane C. Oral sex and the transmission of non-viral STIs. Sex Transm Infect 1998;74:95–100.
2. Coutee F, Trottier AM, Ghattas G, et al. Risk factors for Oral Human Papillomavirus in Adults Infected and not Infected with Human Immunodeficiency Virus. Sex Transm Dis 1997;24:23–31.
3. Evans BA, McCormack SM, Kell PD, Parry JV, Bond RA, MacRae KD. Trends in female sexual behaviour and sexually transmitted diseases in London, 1982–1992. Genitourin Med 1995;71:286–290.
4. Wong ML, Chan RKW, Koh D, Wee S. Increase in oral sex and pharyngeal gonorrhea: an unintended effect of a successful condom promotion program for vaginal sex. AIDS 1999;13:1981–1982.
5. Connor CCO, Berry G, Rohrsheim R, Donovan B. Sexual health and use of condoms among local and international sex workers in Sydney. Genitourin Med 1996;72:47–51.
6. Ford K, Wirawan DN, Fajans P, Thorpe L. AIDS knowledge, risk behaviours and factors related to condom use among male commercial sex workers and male tourist clients in Bali, Indonesia. AIDS 1995;9:751–759.
7. Fajans P, Wirawan DN, Ford K. STD knowledge and behaviours among clients of female sex workers in Bali, Indonesia. AIDS CARE 1994;6:459–475.
8. Wong ML, Chan Roy KW, Koh D. A sustainable behavioural intervention to increase condom use and reduce gonorrhoea among sex workers in Singapore: 2-Year Follow-up. Prevent Med 1998;27:891–900.
9. Green LW, Kreuter M. Health Promotion Planning: An Educational and Environmental Approach. Palo Alto: Mayfield Publishers, 1991:125–149.
10. Wong ML, Tan TC, Ho ML, et al. AIDS/HIV-related knowledge and sexual behavior among female sex workers in Singapore. Health Ed J 1994;53:155–162.
11. Strecher VJ, Rosentock IM. The Health Belief Model. In: Glanz K, et al., eds. Health Behaviour and Health Education: Theory, Research and Practice. San Francisco, CA: Jossey-Bass, 1997:41–59.
12. Bandura A. Social foundations of Thought and Action: A Social Cognitive Theory. Englewood Cliffs, NJ: Prentice Hall, 1986.
13. Lee J. Odds ratio or relative risk for cross-sectional data. Int J Epidemiol 1994;23:201–203.
14. 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.
15. Jones DL, Irwin KL, Inciardi J, et al. The high-risk sexual practices of crack-smoking sex workers recruited from the streets of three American cities. Sex Transm Dis 1998:25:187–193.
16. Alary M, Worm AM, Kvinesdal B. Risk behaviours for HIV infection and sexually transmitted diseases among female sex workers from Copenhagen. Int J STD AIDS 1994;5:365–367.
17. Liu IT, So Reynalda S. Knowledge, attitude and preventive practice survey regarding AIDS comparing registered to freelance commercial sex workers in Iloilo City, Philippines. Southeast Asian J Trop Med Public Health 1996;27:696–702.