MALE CLIENTS OF FEMALE SEX WORKERS (FSWs) are an important group vulnerable to HIV infection. They are at high risk for contracting HIV, and they may become a “bridge population” for transmitting the virus to lower-risk populations. However, they are less often than FSWs the targets of HIV/AIDS-related research and preventive interventions. 1 Previous studies targeting male clients of FSWs usually revealed a high prevalence of inconsistent condom use and sexually transmitted diseases (STDs). 2–7
Despite their high risk, male clients of FSWs in Chinese societies have rarely been studied. This may be due to the belief that this population is hard to reach. A survey of new clients attending some social hygiene clinics in Hong Kong showed that more than 80% of the male attendees had patronized FSWs and that more than half did not use condoms. 8 However, the study may have been subjected to a strong selection bias.
The HIV prevalence in Hong Kong has been low. As of the end of December 2002, there were 2015 HIV/AIDS cases officially reported. 9 However, rising trends in the incidences of STD and HIV were observed in mainland China over the past few years. 10–12 It was estimated that the number of adult HIV cases would approach 5 million in mainland China by the year 2005. 12 Cross-border commercial sex activities have become prevalent among male residents of Hong Kong. 13,14 It was found that inconsistent condom use was common among these cross-border commercial sex clients. 14
No HIV/AIDS Behavioral Surveillance Surveys (BSS) has been conducted among male clients of the FSW population in Hong Kong. Only a few reports of BSS in other countries have been reported. 15–17 BSS are systematic, repeated, cross-sectional surveys of HIV/AIDS- and STD-related behaviors among different populations that are at risk of HIV infection. 18 These studies monitor the level of risk behaviors and their changes over time. They therefore provide timely and important information that can be used for such purposes as risk management, national-level program evaluation, and research formulation. 18 It has been suggested that the second generation of HIV/AIDS surveillance should place more emphasis on the monitoring of behavioral trends. 19 Commonly used BSS indicators include prevalence of commercial sex practice, condom use, STDs, same-sex behaviors, and number of sex partners. 15,20
The objective of this study was to conduct BSS among the male clients of FSW population in Hong Kong. The trends of HIV/AIDS-related behaviors and attitudes among male clients of FSWs, as obtained from three BSS samples during the years 1998 to 2001, were compared. Factors that are associated with condom use and STD were also examined.
Hong Kong Chinese male residents aged 18 to 60 years were studied. Three population-based telephone surveys were conducted in July 1998, November 1999, and June 2001. The effective sample sizes for the respective surveys were 1020, 2074, and 2065 (a total of 5159 respondents). The study population comprised all Hong Kong Chinese residents aged 18 to 60 years who were male clients of FSWs in the previous 6 months. These respondents were identified by means of certain screening questions; 143, 252, and 223 such respondents to the 3 surveys completed the questionnaire.
An identical method was used for all three surveys. Telephone numbers were randomly selected from up-to-date residential phone directories, stratified by the three major geographic regions of Hong Kong. Almost 100% of Hong Kong residents have telephones at home (personal communication, Hong Kong Office of the Telecommunications Authority, January 28, 2000). Of each sampled household, a male member aged 18 to 60 years whose past birthday was closest to the day of the interview was invited to participate in the study. The interviews were conducted between 6:00 pm and 10:30 pm. All interviews were performed by trained interviewers. An unanswered call was attempted at least three times before the telephone number was classified as invalid.
After briefing the individual who answered the phone that the study was going to collect opinions about HIV/AIDS prevention, that it was sponsored by the Hong Kong Council for the AIDS Trust Fund, and that participation would help the government formulate policy, the interviewer invited the male household member aged 18 to 60 years whose past birthday was closest to the day of the interview to participate in the study. Approximately 77% of the households with a male member had only one male member, 17.9% had two, and 5% had more than two (personal communication, Department of Census & Statistics, April 26, 2002).
After a respondent was recruited, the interviewer first asked a few less sensitive questions to establish rapport (part I). This was important to facilitate the subsequent asking of sensitive questions (part II). The respondents were told that the second part of the interview would cover questions on HIV/AIDS-related behaviors and that because these questions were prerecorded in a computerized phone system (the “dot-line” service), they would need only to key in their responses. Respondents were guaranteed that their telephone number would not be recorded. Those who agreed to participate in the second part of the interview were connected to the dot-line service via the conference line service (which allows one of the two parties to call a third one without hanging up and allows all three parties to have a simultaneous phone conversation). The interviewer left the line after the connection was complete.
The respondents then keyed in their responses after listening to the prerecorded questions. The dot-line telephone number was not released to the respondents. The service has often been used by television stations for public opinion polling and is widely known to the public. Anonymity and confidentiality were therefore enhanced. Two previous studies showed that the interactive computerized call-in method yielded high completion rates of 87% and 97%. 14,21 The methodological advantages of this approach (e.g., minimizing reporting bias) have been documented, 22–23 and it has been used in a number of studies. 24–26
Of all valid household contacts (with at least one Chinese male aged 18–60 years; n = 2213, 4713, and 4613 in the 3 surveys), approximately 50% of them had one eligible male member participate in the interview and complete the nonsensitive part I (51.6, 46.8, and 48.2%, respectively); 89.2%, 94.1%, and 92.9% of these respondents went on to complete the sensitive part II. A fairly high percentage (>40%) of the incomplete part II questionnaires were caused by technical problems (e.g., use of a pulse-mode phone). The overall response rates—defined as the number of completed interviews, divided by the total number of valid household contacts—were 46.1% (1020/2213), 44.0% (2074/4713), and 44.8% (2065/4613) in 1998, 1999, and 2001, respectively.
The same structured questionnaire was used in all three surveys. Data on the following BSS indicators, including behavioral and attitudinal ones, were collected. Behavioral indicators included (1) patronage of FSWs in the 6 months before the interview; (2) consistency of condom use (i.e., used condom every time) during sexual intercourse with FSWs in the past 6 months; (3) self-reported STD in the past 6 months; (4) number of female sex partners in the past 6 months; (5) HIV antibody testing in the past 6 months (asked only in 1999 and 2001); and (6) the geographic locations where commercial sex transactions took place. Attitudinal indicators included (1) self-perceived chance of contracting HIV in the future and (2) perceived efficacy of condom use for HIV prevention.
Part I of the questionnaire inquired about age, education level, and level of HIV-related knowledge (respondents were asked to name three routes of HIV transmission). The above-mentioned question on perceived efficacy of condom use was asked in part I, and questions related to the rest of the BSS indicators were asked in part II of the questionnaire.
The surveillance data were analyzed at three levels: those of the general adult male population, those from the male FSW client population, and those from subsets of this male FSW client population who had or had not purchased sex across the border between Hong Kong and mainland China. For each of the three levels of analysis, the percentage distributions of the studied variables were tabulated by the year of survey. To examine between-year differences, the chi-square test for trend and logistic regression analyses (with the 1998 survey as the reference group and with adjustment for age, education level, and an indication of HIV-related knowledge attained) were performed. In addition, stepwise multivariate logistic regression analyses were applied to the male FSW client sample to examine factors that were associated with consistent condom use with FSWs and the prevalence of self-reported STD. Candidate variables initially considered included calendar year of survey, age, education level, knowledge about HIV transmission routes, having had four or more female sex partners in the previous 6 months, self-perceived susceptibility to HIV infection, perceived efficacy of condom use for HIV prevention, having patronized FSWs in mainland China in the past 6 months, having visited two or more geographic locations for commercial sex, and having undergone HIV antibody testing in the past 6 months. The subsets of univariately significant variables were entered into the above-mentioned stepwise logistic regression analyses. Relevant odds ratios (ORs) and their respective 95% confidence intervals (CIs) were presented. A P value <0.05 was considered statistically significant. All statistical analyses were conducted with SPSS for Windows, version 11.0.
BSS for the General Adult Male Population
The demographic backgrounds and HIV-related knowledge of the respondents are summarized in Table 1. The distributions of these characteristics and responses were more or less comparable across the years. The differences were of statistical significance, however, because the sample sizes were relatively large (Table 1). These differences (age, education level, and HIV-related knowledge) were adjusted in the subsequent analyses.
Among all respondents, the percentage of the general adult male population patronizing a FSW in the previous 6 months decreased from 14% (95% CI: 11.9–16.1) in 1998 to 12.2% (95% CI: 10.8–13.6) in 1999 and 10.8% (95% CI: 9.5–12.1) in 2001 (chi-square test for trend, P < 0.01;Table 2). Around 1.5% (95% CI: 0.8–2.2%) of the entire general adult male population had contracted an STD in the previous 6 months, and no significant time trend was observed (chi-square test for trend, P = 0.78). The percentages of respondents having had multiple female sex partners (≥2) in the past 6 months were about 10% to 11%. No statistically significant changes over time were observed (chi-square test for trend, P = 0.20). Around 7% to 8% of the general male population had undergone HIV antibody testing in the past 6 months (data collected only in the 1999 and the 2001 samples; chi-square test for trend, P = 0.17).
No apparent sample differences were observed for the perceived susceptibility of contracting HIV in the future (chi-square test for trend, P = 0.77). Close to 70% of the general adult male population perceived no chance of contracting HIV. Similarly, no significant differences were observed in the perceived efficacy of condom use for HIV prevention (Table 2).
The statistical significance of all adjusted odds ratios (adjusted for age, education level, and HIV-related knowledge) was consistent with that of the above-mentioned chi-square test values for trends (Table 2).
BSS for the FSW Client Population
A total of 618 male clients of FSWs were identified by the study (n = 143, 252, and 223, respectively, for the 3 samples;Table 3). There were no significant between-year differences in the percentages of consistent condom use with FSWs in the past 6 months (72.7%, 74.6%, and 77.1%; chi-square test for trend, P = 0.33;Table 3), in the percentages reporting having contracted an STD (5.6%, 6.4%, and 4.0%; P = 0.44), in the percentages having tested for HIV antibody (16.4% and 16.1% in the year 1999 and 2001; P = 0.94), in the percentages of perceived susceptibility of HIV infection in the future (53.8%, 55.8%, and 52.9%; P = 0.80), and in the percentages perceiving condom use as having high efficacy for HIV prevention (67.8%, 67.6%, and 72.1%; P = 0.34). The percentages of respondents having had four or more female sex partners in the past 6 months increased from 18.9% in 1998 to 30.9% in 2001 (chi-square test for trend, P < 0.01;Table 3).
The percentages of respondents having patronized FSWs in mainland China in the past 6 months were 47.6%, 63.1%, and 53.8% for the three study years (chi-square test for trend, P = 0.45), and the percentages having patronized FSWs in more than one geographic location were 52.5%, 61.4%, and 53.6% for the respective 3 years (P = 0.89, Table 3).
The statistical significance of all adjusted odds ratios (adjusted for age, education level, and HIV-related knowledge) was consistent with that of the above-mentioned chi-square test for trends, except for the case of patronage of FSWs in mainland China in the previous 6 months. In this case, male FSW clients in 1999 were more likely to have patronized FSWs in mainland China than were clients in 1998 (Table 3).
BSS for FSW Clients Having or Not Having Purchased Sex in Mainland China
Respondents who were FSW clients were further stratified by whether they had patronized FSWs in mainland China (Table 4). For the group of cross-border FSW clients, there were no between-year differences in the consistency of condom use, prevalence of self-reported STD, prevalence of HIV antibody testing, perceived chance of contracting HIV in the future, and perceived efficacy of condom use for HIV prevention (Table 4). The prevalence of having had four or more female sex partners among the cross-border FSW clients increased over the years (chi-square test for trend, P = 0.04). However, this was not statistically significant after adjustment for age, education level, and HIV-related knowledge (Table 4). For those who had patronized FSWs in places other than mainland China, no significant between-year differences were observed for the six BSS indicators (Table 4).
Except for the above-mentioned associations with having had four or more female sex partners, the statistical significance of all adjusted odds ratios (adjusted for age, education level, and HIV-related knowledge) was consistent with that of the above-mentioned findings of the chi-square test for trends (Table 4).
In addition, FSW clients who had purchased sex in mainland China, as compared with those who had not, were less likely to be consistent condom users and more likely to have contracted an STD, had a greater number of female sex partners in the previous 6 months, and perceived some chances of contracting HIV in the future, after adjustment for calendar year of survey, age, education level, and HIV-related knowledge (Table 4). No significant differences over time in the prevalence of HIV antibody testing and perceived efficacy of condom use for HIV prevention were observed for such FSW clients (Table 4).
Factors Associated with Consistent Condom Use with FSWs Among Male Clients of FSWs
Univariately, older age, lower education level, self-reported STD in the past 6 months, four or more female sex partners in the past 6 months, patronizing FSWs in mainland China in the past 6 months, and visiting two or more geographic locations for commercial sex in the past 6 months were significantly associated with a low likelihood of consistent condom use (P < 0.05, chi-square test;Table 5). Other variables (calendar year of survey, knowledge about HIV transmission routes, self-perceived susceptibility to HIV infection, perceived efficacy of condom use for HIV prevention, and HIV antibody testing) were not of statistical significance in the univariate analysis (P > 0.05).
Results of a stepwise multivariate logistic regression analysis using univariately significant variables as inputs to predict consistency of condom use during sexual intercourse with FSWs in the past 6 months showed that respondents who had attained a higher level of education (OR for ≥ secondary grade 6 versus ≤ secondary grade 4 = 2.43; 95% CI: 1.35–4.34; P < 0.01) were more likely to be consistent condom users. Respondents who had had four or more female sex partners in the past 6 months (OR for yes versus no = 0.33; 95% CI: 0.21–0.52; P < 0.001), had contracted STD in the past 6 months (OR for yes versus no = 0.21; 95% CI: 0.09–0.5; P < 0.001), and had patronized FSWs in mainland China in the past 6 months (OR for yes versus no = 0.45; 95% CI: 0.28–0.72; P < 0.01) were less likely to be consistent condom users (Table 5). Having visited two or more geographic locations for commercial sex and age were significant in the univariate analysis but not in the multivariate analysis (Table 5).
Factors Associated with Self-Reported STD Prevalence Among Male Clients of FSWs
Univariately, inconsistent condom use with FSWs in the past 6 months, having had four or more female sex partners in the past 6 months, having patronized FSWs in mainland China in the past 6 months, and having undergone an HIV antibody test in the past 6 months were significantly associated with a higher likelihood of contracting an STD in the past 6 months (P < 0.05, chi-square test). Other variables, including calendar year of survey, age, education level, knowledge about HIV transmission routes, self-perceived susceptibility to HIV infection, perceived efficacy of condom use for HIV prevention, and having visited two or more geographic locations for commercial sex, were not of statistical significance in the univariate analysis (P > 0.05).
A stepwise multivariate logistic regression analysis was performed with use of the three aforementioned univariately significant variables as inputs to predict self-reported STD in the past 6 months (although HIV antibody testing also was univariately significant, it was not included in the stepwise model because the question was asked only in the 1999 and 2001 surveys). Results showed that respondents who had had four or more female sex partners in the past 6 months (OR for yes versus no = 2.75; 95% CI: 1.28–5.89; P < 0.01) and had patronized FSWs in mainland China in the past 6 months (OR for yes versus no = 5.36; 95% CI: 1.58–18.22; P < 0.01) were more likely to have contracted STD in the past 6 months. Those who had always used condoms during sexual intercourse with FSWs (OR for yes versus no = 0.20; 95% CI: 0.09–0.46; P < 0.001) were less likely to have contracted STD in the past 6 months.
This BSS provides stakeholders with a set of relevant BSS indicators and data obtained with a special collection method. 22–23 A reasonably large number of male FSW clients were recruited by this method. It is one of the very few population-based studies of FSW clients conducted in the region.
By carrying out regular cross-sectional surveys and strictly using the same methodology and questions, we attempted to capture any changes in risk behaviors in an at-risk population, in this case the FSW client population. One of the purposes for carrying out a BSS is to evaluate the overall effectiveness of the relevant prevention programs. 20 The overall effectiveness of global efforts rather than individual programs was assessed by analysis of the BSS data. Because there is no apparent improvement in the reduction of risk behaviors (as reflected by the relevant BSS data), despite the large amount of resources spent on corresponding prevention programs, the effectiveness and the coverage of these programs is questionable. In other words, among FSW clients in Hong Kong, risk behaviors (such as inconsistent condom use, multiple sex partners) have been common and have not been reduced over the study period. The results are consistent with those obtained in another study of males who crossed the border to have sex with FSWs in mainland China. 27 These findings could be due to the ineffectiveness of existing programs or the low coverage of the programs. The latter, at least, may be true of Hong Kong. In fact, there are very few such prevention programs in Hong Kong, and only a few workers are responsible for the prevention work in this population.
By applying the surveillance data of this study to the census data, the local size of the male FSW client population who had had sexual intercourse with FSWs in the past 6 months was estimated to be in the range of about 0.23 to 0.3 million, which is much larger than the estimated size of the population of men who have sex with men in Hong Kong (about 0.04 million). 26 The current resource allocation system and nongovernmental organizations do not seem to take into account the size and levels of risk behavior of the target populations. The prevention work among commercial sex clients in Hong Kong is underdeveloped. The low intensity of community and governmental responses is incomprehensible in view of the relatively resourceful economic situation in Hong Kong and the wide recognition of the importance of the problem. 28,29
An equally important message is that there is a big differential in the level of risk between the FSW clients having or not having patronized FSWs in mainland China. It was found that the former group was much more likely to be inconsistent condom users; they also were more likely to have contracted STD, to have had four or more female sex partners in the past 6 months, and to perceive some chance of contracting HIV. Therefore, these cross-border FSWs clients are serving as a “bridge population” that transmits HIV across geographic areas and across genders (because they seldom use condoms with their regular sex partners 14). The Guangdong Province, which is adjacent to Hong Kong, ranked fourth among all provinces in China in terms of the number of reported HIV-positive cases. 30 There are warnings that the HIV epidemic will explode in mainland China, and southern China would be one of the more strongly affected areas. This makes the future prevalence of HIV in Hong Kong uncertain. It also suggests that it would be much more cost-effective to give higher priority to prevention work among cross-border FSW clients than among those who patronize FSWs in Hong Kong. However, it was recently announced by the Hong Kong Council for the AIDS Trust Fund that it would not consider funding for cross-border projects in mainland China targeting Hong Kong residents or to educate FSWs whose clients were mostly Hong Kong males (about 67%31). The funding policy is therefore incompatible with the size of the problem and the suggestions made by the local Community Planning Committee 29 and the Hong Kong Advisory Council on AIDS. 32,33
Similar to the results of another study of cross-border FSW clients, 27 the prevalence of HIV antibody testing among the male clients of FSWs in our study was low (about 16%). Voluntary counseling and testing (VCT) has been demonstrated to be an effective means of HIV prevention. 34 Promotion of HIV antibody testing of Hong Kong FSW clients is therefore warranted. The low testing rate also suggests that the current HIV prevalence, which is determined primarily on the basis of reported data, may be underestimated. Unfortunately, no biologic (seroprevalence) surveillance has been carried out among this group in Hong Kong, and such surveillance needs to be strengthened.
The proportion of the general adult male population who had patronized FSWs decreased by 3.2% (P < 0.01) over the study period. A similar decrease in the proportion of cross-border travelers having patronized FSWs in mainland China was also observed in another study. 27 It is possible that the very drastic economic downturn is at least partially responsible for the decrease. The unemployment rate went up from 2.2% in 1997 35 to 7.2% in December 2002, 36 and real estate prices dropped by >60% during the study period. 37 It is not certain at this stage whether the observed decrease in the number of FSW clients constitutes a secular trend. The question might be answerable with future rounds of BSS.
It was also noted that among those who visited FSWs, the average number of female sex partners increased in the 2001 study. The number of sex partners was used as a proxy for the frequency of FSW visits and the number of FSWs contacted, which were important variables that had not been measured because the sensitive questions asked in the study were limited. A lengthy questionnaire would further lower the response rate. One speculation is that the increase was related to the aforementioned decrease in the number of FSW clients in 2001. It is possible that those who did visit FSWs in 1998 but not in 2001 were occasional FSW patrons, and the rest of the FSW clients who still visited FSWs in 2001 might thus have had more sex partners. Again, future BSS would indicate whether a secular trend exists.
In sum, the risk behaviors of the FSW client group, a sizable risk group in Hong Kong, and especially of those who purchased sex across the border will largely determine the future HIV prevalence in Hong Kong. With the low testing rate and the lack of an effective seroprevalence surveillance system, the degree of underreporting of HIV prevalence in Hong Kong remains uncertain. The BSS results did not show significant signs of reduction of risk behaviors in the FSW client group. Commercial sex behavior across the border is worth investigation, and a regional approach is needed. The program coverage of the FSW client population has been extremely low. Hong Kong seems unprepared to face the impact of an emerging epidemic in mainland China, which will easily be bridged over to Hong Kong via the FSW client population.
The study had a number of limitations. For sensitive questions such as patronage of FSWs, HIV antibody testing, and self-reported STD incidence, response bias in the direction of underreporting is expected. The study used an innovative population-based data collection method to recruit FSW client respondents and to capture the extent of their risk behaviors. Such a method has been shown to reveal higher levels of risk behaviors, in comparison with conventional telephone survey methods. 22–23 With BSS, which involve use of the same methodology and the same questionnaire, the effects of response bias is rather constant in different surveys. The constancy of the biases is probably as important as their magnitude when the purpose is to detect changes. The stability of the data in the three surveys for most of the questions, as a result of the BSS design, illustrates that the size of the biases should also have been stable. Furthermore, in an independent study, 27 the prevalence of inconsistent condom use among those who had had sexual intercourse with FSWs in mainland China in the preceding trip to mainland China was around 39%, 37%, and 39.2% in 1998, 1999, and 2001, versus similar results of 41.2%, 30.8%, and 30% in 1998, 1999, and 2001, as observed in this study. This validates the results of the current study.
Selection bias is another serious issue that may have affected the results of the study, especially because the response rates of the surveys were below 50%; other less sensitive local telephone surveys usually have a response rate of around 50% to 60%. 38–40 Population-based sampling of male FSW clients is always difficult because there is no sampling frame to work with. Many studies, in fact, are venue-based; other forms of selection biases exist in these studies. Random sampling is also extremely difficult under those circumstances. This explains why population-based studies employing random sampling are so scarce. Another possible source of selection bias may be related to follow-ups of noncontact households. In this study, three follow-up calls were made at different hours and days before a household was declared invalid. Some of these households might in fact contain males who would have been eligible for the study. The three-calls rule has been used in a number of other local studies. 23,25,26,41,42 Again, as the same methodology was used, the magnitude of the biases due to selection should be comparable for the three surveys, and changes would represent changes in behaviors rather than in the size and nature of biases. The comparison of the cross-border data with those obtained from another independent study 27 also supported the validity of the data collected in this study.
There are other limitations. For instance, although most of the households in Hong Kong have telephones, not all phone numbers are listed in the telephone directory. There are no data on the percentage of listings, but the figure should be quite high; when a new telephone number is issued, it is automatically listed (no application or fee is necessary) unless the applicant requests otherwise. Furthermore, the scope and depth of the questions were limited by the length of the questionnaire. Because the questions were of a sensitive nature, only the key indicator questions were asked. Therefore, questions such as frequency of FSW visits and number of FSWs visited were not asked, to ensure that the response rates to the part II questionnaire (which were very high: 89.2–94.1%) would be high enough to prevent other selection bias.
Despite the above-mentioned limitations, the study is a step forward in the investigation of HIV-related risk behaviors. With the BSS design, we utilized a special data collection methodology and, more important, conducted a series of population-based studies of a random-sampling design and successfully identified a reasonably large number of FSW clients. This design may form the basis of similar studies in other countries.
1. Ghys PD, Jenkins C, Pisani E. HIV surveillance among female sex workers. AIDS 2001; 15( suppl 3): S33–S40.
2. Estebanez P, Rua-Figueroa M, Aguilar MD, Bru F, Rog E, Zunzunegui MV. Condom use in clients of prostitutes in Spain. AIDS 1996; 10: 445–446.
3. Fajans P, Ford K, Wirawan DN. AIDS knowledge and risk behaviors among domestic clients of female sex workers in Bali, Indonesia. Soc Sci Med 1995; 41: 409–417.
4. 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.
5. Wong KH, Lee SS, Lo YC, Lo KK. Condom use among female commercial sex workers and male clients in Hong Kong. Int J STD AIDS 1994; 5: 287–289.
6. Tabrizi SN, Skov S, Chandeying V, Norpech J, Garland SM. Prevalence of sexually transmitted infections among clients of female commercial sex workers in Thailand. Sex Transm Dis 2000; 27: 358–362.
7. Espirito Santo ME, Etheredge GD. HIV prevalence and sexual behaviour of male clients of brothels’ prostitutes in Dakar, Senegal. AIDS Care 2003; 15: 53–62.
8. Department of Health. Sentinel behavioural surveillance of clients attending government sexually transmitted disease clinics. Hong Kong STD/AIDS Update 2000; 6: 8.
9. Hong Kong Advisory Council on AIDS. Reported HIV/AIDS statistics. ACA Newsfile 2003; 10: 6.
10. Chen XS, Gong XD, Liang GJ, Zhang GC. Epidemiologic trends of sexually transmitted diseases in China. Sex Transm Dis 2000; 27: 138–142.
11. Chen L, Feng TJ, Li LC, He JF. Analysis on the trends of HIV/AIDS epidemic in Shenzhen. Chin J Sex Transm Infect 2002; 2( 2): 16–20.
12. UNCHINA [Joint United Nations Programme on HIV/AIDS, China Office]. Perspective of the UNAIDS China office on the situation of HIV/AIDS in China in 2000. Retrieved 19 June 2002 from http://www.unchina.org/unaids/ekey3right1.html
13. Key SW, DeNoon DJ, Boyles S. Cross-border sex raises AIDS alarm in Hong Kong. AIDS Weekly Plus 1997; Sept 29: 15.
14. Lau JTF, Thomas J. Risk behaviours of Hong Kong male residents travelling to mainland China: a potential bridge population for HIV infection. AIDS Care 2001; 13: 71–81.
15. Mills S, Benjarattanaporn P, Bennett A, et al. HIV risk behavioral surveillance in Bangkok, Thailand: sexual behavior trends among eight population groups. AIDS 1997; 11( Suppl 1): S43–S51.
16. Utomo B, Dharmaputra NG, Haryanto B, Harono D, Makalew R. Indonesia HIV/AIDS Behavioral Surveillance Survey: Results from the Cities of North Jakarta, Surabaya, and Manado. Jakarta: Center for Health Research, 1997.
17. Jenkins C, Rahman H. Rapidly changing conditions in the brothels of Bangladesh: impact on HIV/STD. AIDS Edu Prev 2002; 14( 3 Suppl A): 97–106.
18. Mills S, Saidel T, Bennett A, et al. HIV risk behavioral surveillance: a methodology for monitoring behavioral trends. AIDS 1998; 12( Suppl 2): S37–S46.
19. UNAIDS/WHO. Guidelines for Second Generation HIV Surveillance. Geneva: UNAIDS/WHO, 2000.
20. Family Health International. Behavioral Surveillance Surveys: Guidelines for Repeated Behavioral Surveys in Populations at Risk of HIV. Arlington, Virginia: Family Health International, 2000.
21. Lau JTF, Thomas J, Lin CK. HIV-related behaviors among voluntary blood donors in Hong Kong. AIDS Care 2002; 14: 481–492.
22. Lau JTF, Thomas J, Liu JLY. Mobile phone and interactive computer interviewing to measure HIV-related risk behaviours: the impacts of data collection methods on research results. AIDS 2000; 14: 1277–1279.
23. Lau JTF, Tsui HY, Wang QS. Effects of two telephone survey methods on the level of reported risk behaviors. Sex Transm Infect (In press).
24. Lau JTF, Wong WS. Behavioural surveillance of sexually-related risk behaviours for the cross-border traveller population in Hong Kong: the evaluation of the overall effectiveness of relevant prevention programmes by comparing the results of two surveillance surveys. Int J STD AIDS 2000; 11: 719–727.
25. Lau JTF, Tang ASY, Siah PC, Tsui HY. Assessment of HIV-related sexual risk behaviors among the general female population in Hong Kong. Arch Sex Behav 2002; 31: 535–542.
26. Lau JTF, Kim JH, Lau M, Tsui HY. HIV-related behaviors and attitudes among Chinese men who have sex with men in Hong Kong: a population-based study. Report to the Hong Kong Council for the AIDS Trust Fund. Hong Kong: Council for the AIDS Trust Fund, August 2002.
27. Lau JTF, Tsui HY. HIV/AIDS behavioral surveillance surveys of the cross-border sex-networker population in Hong Kong from 1997 to 2001. Report to the Hong Kong Council for the AIDS Trust Fund. Hong Kong: Council for the AIDS Trust Fund, August 2002. Sex Transm Dis (in press)
28. Advisory Council on AIDS. Moving Ahead Together: Expanding Hong Kong's Response to AIDS. ACA External Consultancy Review 1998. Hong Kong: Advisory Council on AIDS, 1998.
29. Community Planning Committee. Final Report of the Community Planning Process. Hong Kong: Community Planning Committee, August 2001.
30. The China AIDS/STI Prevention and Control Information Network. Available online at http://www.aids.net.cn
31. Lau JTF, Tsui HY, Siah PC, Zhang KL. A study on female sex workers in southern China (Shenzhen): HIV-related knowledge, condom use and STD history. AIDS Care 2002; 14: 219–233.
32. Advisory Council on AIDS. Recommended HIV/AIDS Strategies for Hong Kong: 2002 to 2006. Hong Kong: Advisory Council on AIDS, May 2002.
33. AIDS Prevention and Care Committee. HIV prevention in cross border travelers: principles of strategy. Hong Kong: Advisory Council on AIDS, March 2001.
34. Merson MH, Dayton JM, O'Reilly K. Effectiveness of HIV prevention interventions in developing countries. AIDS 2000; 14( Suppl 2): S68–S84.
35. Hong Kong Census and Statistics Department. Hong Kong Annual Digest of Statistics. Hong Kong: Government Printer, 2002.
38. Lai G, Wong O. The tie effect on information dissemination: the spread of a commercial rumor in Hong Kong. Social Networks 2002; 24: 49–75.
39. Kwan AC, Hu WH, Chan YK, Yeung YW, Lai TS, Yuen H. Prevalence of irritable bowel syndrome in Hong Kong. J Gastroenterol Hepatol 2002; 17: 1180–1186.
40. Chou KL, Mak KY, Chung PK, Ho K. Attitudes towards mental patients in Hong Kong. Int J Soc Psychiatry 1996; 42: 213–219.
41. Tang CS. The rate of physical child abuse in Chinese families: a community survey in Hong Kong. Child Abuse Negl 1998; 22: 381–391.
42. Lam CL, Fong DY, Lauder IJ, Lam TP. The effect of health-related quality of life (HRQOL) on health service utilisation of a Chinese population. Soc Sci Med 2002; 55: 1635–1646.