Lau, Joseph T. F. PhD*; Wang, Renfan MD†; Chen, Hongyao MD†; Gu, Jing MSc*; Zhang, Jianxin MD, MSc‡; Cheng, Feng PhD§; Zhang, Yun MD∥; Zhang, Linglin MD, MSc¶; Tsui, Hi-Yi MPhil*; Wang, Ning PhD#; Lei, Zhangquan MD†; Zhong, Xiaodong MD†; Lan, Yajia PhD‡
IT IS OFFICIALLY ESTIMATED that as of December 2005, there were about 0.65 million people living with HIV/AIDS in China.1 International organizations have joined force with the Chinese workers to promote HIV-related prevention and care in China2 and evaluation of such efforts is greatly warranted. Evaluation of the overall effectiveness of programs targeting HIV prevention in a community, such as those promoting condom use among female sex workers (FSWs), have been reported in different countries.3–8
Behavioral Surveillance Surveys (BSS), consisting of a series of periodical cross-sectional surveys using the same methodology,9 can be used to assess the overall effectiveness of programming efforts at a community level.10 BSS has been implemented in a number of countries, such as the United States, Thailand, Vietnam, INonesia, Nigeria.11–15
Since 2000, the China-UK HIV/AIDS Prevention and Care Project (China-UK Project) has implemented HIV/AIDS prevention and care projects in 83 counties in the Sichuan and Yunnan provinces. As of December 2005, it has spent about 20 million pounds, and 54,000 people in Sichuan had been served by 90 interventions. The Project was managed by its National and Provincial Management Offices, and interventions were implemented at the county level. The Project also supports the development of BSS in the Sichuan and Yunnan provinces in China.16 Both funding and technical support have been provided by United Kingdom’s Department for International Development and Family Health International (FHI). The first round of BSS was implemented in 2002, with 8 sites in Sichuan and 3 sites in Yunnan; the number of sites then increased to 21 and 11, respectively, in 2003 and 2004, and 21 and 24, respectively, in 2005. Community-based samples of FSWs (i.e., respondents were not recruited from institutions such as retention centers nor were they recruited from clinics) were recruited in these BSS studies.
The HIV epidemic in China has been driven by the high prevalence in the injection drug user (IDU) population.17 There are concerns that HIV would eventually be spread to the general population through various bridge populations.18–20 Heterosexual transmissions comprised 19.8% of the country’s estimated HIV cases in 2003, which increased to 36.3% in 2005.21 There are 4 to 6 million FSWs in China22 and HIV and sexually transmitted diseases (STD) prevalence in this group respectively ranged from 1.4% to 6.7%23,24 and 10% to 60%.25,26
In 2005, Sichuan had a population size of about 87.3 million27 and ranked sixth among all Chinese provinces in the number of reported HIV cases.28 The HIV prevalence among IDUs in different surveillance sites in Sichuan ranged from 0% to 66.0% in year 2003.29 Since 2003, the China-UK Project has implemented intervention programs (Expanded Comprehensive Integrated Response, ECIR) in a number of cities in Sichuan, combining BSS with various types of interventions.30
Dazhou city of Sichuan has a total population of about 6.5 million. It is divided into 7 counties, 2 of which are DC and TC, which respectively have population sizes of 1.06 million and 0.38 million, and are only about 60 km apart. A total of 516 and 18 HIV cases had been reported in DC and TC respectively during 2003 to 2005 [personal communication with director of Dazhou Center for Disease Control and Prevention (CDC), Sichuan]. Both counties are served by the Dazhou CDC, and annual BSS were conducted in the urban areas of these 2 counties in the years 2003, 2004, and 2005 by the same team of workers. During these 3 years, the ECIR of the China-UK Project was implemented in DC but not in TC.
Before the implementation of the China-UK Project, HIV intervention in Dazhou was only limited to publicity education activities such as those organized on Word AIDS Days. The China-UK Project’s ECIR adopted an integrated approach. High level government support was sought, and a HIV prevention committee was chaired by the former mayor of Dazhou to facilitate interdepartmental coordination; the CDC also set up project offices at city and county levels. As a result, 1 outreach team, 10 voluntary counseling and testing sites, 1 needle exchange centre, 1 methadone maintenance treatment centre, 7 community-based HIV counseling centers/clinics, 2 STD clinics, and 7 community-based support groups were set up in DC. Training was provided to 26 CDC staff, 316 HIV/STD clinicians, 54 interviewers, 245 officials, and 427 policemen, etc. Intervention activities targeting FSWs include mobilization of community resources (e.g., family planning workers and officials), seminars and training for keepers of entertainment establishments and FSWs, publicity events and advertisements via different mass media, outreach education provided by 20 trained peer educators, voluntary counseling and testing services, free STD checkup and support services, and social marketing of condoms, etc. DC is therefore regarded as the intervention community and TC as the control community. During the study period, there were no other nongovernmental organizations or other major intervention projects on HIV prevention serving FSWs in DC.
This study therefore, documented and compared the results of the annual BSS on FSWs during 2003 to 2005 in DC and TC. The approach of using paired communities to evaluate HIV prevention efforts have been reported in other countries.4,6 The changes in the evaluative behavioral parameters in both sites were compared so as to evaluate overall effectiveness of the ECIR of the China-UK Project implemented in DC. Factors associated with condom use were also investigated.
Materials and Methods
The methodology of the BSS in Sichuan (including Dazhou) has been documented in detail elsewhere.29 Supported by international and national experts, a series of meetings were held in 2002 to translate and to modify the BSS questionnaires developed by the FHI.
Intensive training sessions (3-day workshops) were provided to interviewers by experienced researchers from the School of Public Health of Sichuan University, staff of National and Provincial CDC. CDC core staff also attended several training workshops organized by the China-UK Project. A detailed operational manual was prepared and used. Interviewers kept a log book describing their fieldwork and were supervised by senior CDC staff who performed field and office editing. Data collection was also monitored by the provincial and national level experts. An independent international monitoring team was sent to the sites to review the documentation and quality of the field work. The team also visited some establishments and interviewed some respondents and was satisfied with the quality of the study.
Mapping exercises were done for sex work establishments annually before the implementation of the BSS. A number of establishments (about 30) were randomly selected and gatekeepers of these establishments were contacted by the CDC staff. The sample size was about 400 per survey (200 in TC and 200 in DC), which was recommended by the national guideline.31 In TC and DC, the number of FSWs to be recruited from an establishment was roughly proportionate to the estimated number of FSWs working in that establishment. Fieldworkers visited these establishments at different time slots and the convenience sampling method was used to recruit study respondents. Pilot tests were conducted, and the questionnaire and procedures were then refined by an expert panel. Face-to-face interviews, using an anonymous structured questionnaire, were administered. Informed consent was obtained verbally, and data confidentiality was assured. All respondents were recruited from the sampled establishments. The fieldworker signed a form pledging that he or she has explained the information clearly to the respondents. Ethics approval has been obtained from the national CDC in China. A cash incentive of RMB50 (about US$6) was given to the respondent.
Respondents were asked questions about their background characteristics (listed in Table 1). The parameters for evaluation included HIV-related knowledge and perceptions (listed in Table 2), whether condoms were used with sex work clients (in the last episode and consistently during the last month) and with regular sex partners (in the last episode and consistently during the last month), self-reporting of ever having STD symptoms, and injecting drug use behaviors in the last 12 months. Health services utilization variables including whether ever having received voluntary HIV antibody testing, STD checkups in the last 6 months, and whether having used any other HIV-related preventive services in the last 12 months.
The distributions of the respondents’ background characteristics were tabulated by the year of the study. Time trends were tested by χ2 test for trends, and within each year, the differences of the respondents’ characteristics in DC versus TC were tested by Pearson χ2 test. The prevalence of the parameters for evaluation (e.g., condom use with sex work clients and regular sex partners) in DC and TC in respective years was tabulated, and the differences were compared by odds ratios obtained from multivariate logistic regression methods, adjusting for age, education level, type of FSW, and duration of sex work. Factors predicting some of the evaluative measures were investigated using univariate odds ratios; univariately significant variables were then used as candidate variables for multivariate stepwise logistic regression analyses. SPSS version 13.0 was used for data analysis; P <0.05 was considered statistically significant.
Background Characteristics of Respondents
A number of the respondents’ characteristics in the 3 surveys were significantly different (Table 1). Within a particular year, many of these characteristics of respondents in DC were also significantly different from those of TC. In particular, the proportion of street workers in 2003 was higher than those of 2004 and 2005. Adjustment of background factors was hence carried out when results obtained from TC and DC were compared in data analyses. Objection from clients is the leading reason for not using a condom in the last episode of sex work, followed by “no need,” “haven’t thought about it,” and other reasons (Table 1).
HIV-Related Knowledge and Perceptions in DC Versus TC
Inspecting the baseline data (2003) of the 5 HIV-related knowledge items and the 4 perception items, the differences between TC and DC were not of statistical significance (expect for “mosquito bites”) (Table 2). In 2004 and 2005, the percentages of appropriate responses to HIV-related knowledge items were significantly higher in several cases in DC when compared with TC. In 2005, 96.1% and 67.2%, respectively, of all respondents in DC versus TC had at least 4 of 5 items appropriately answered (OR comparing DC vs. TC adjusting for background characteristics = 7.9 and 17.3, respectively, for 2004 and 2005, P <0.01). Comparing 2005 and 2003 data, the percentages of respondents not believing that oral or external chemical contraceptive use could prevent HIV increased both in DC and TC (P <0.05, χ2 trend test). Higher percentages were found for all these variables in DC in 2004 and for the variable “douching vagina after sex” in 2004 and 2005, when compared with TC (adjusted OR = 3.3–10.6, P <0.05, Table 2).
Although the prevalence of respondents perceiving themselves to be susceptible to HIV infection in DC and TC did not differ in 2003, respondents in DC felt much less susceptible to HIV infection in 2005 (OR = 0.1, P <0.01, Table 2). The prevalence increased in TC but decreased in DC (P <0.05).
Condom Use Behaviors in DC Versus TC
In the baseline survey, the differences in the prevalence for condom use between DC and TC were not of statistical significance (Table 3): condom use with the last client (DC vs. TC: 82.5% vs. 85.6%), consistent condom use with clients in the last month (50% vs. 37.1%), last-time condom use with regular sex partner (25.2% vs. 21.3%) and last-month condom use with regular sex partner (8.7% vs. 5.0%). During the study period, however, substantial increases in the prevalence of condom use with both sex work clients and regular sex partners (both for the last time and in the last month) were recorded in DC (P <0.05), but not in TC (P >0.05, χ2 test for trend, Table 3). The adjusted odds ratio comparing prevalence of the 4 studied condom use variables in DC versus TC ranged from 2.2 to 33.2 (P <0.05) in 2004 and ranged from 3.8 to 8.3 (P <0.05) in 2005 (Table 3). The association between consistent condom use with regular sex partners and sex worker clients in the last month was statistically significant (Spearman correlation coefficient = 0.348, P <0.001).
Self-Reported STD and IDU Behavior in DC Versus TC
At the baseline (2003), 32.5% and 34.2% of the respondents in DC and TC respectively self-reported that they had ever had STD symptoms (P >0.05). The prevalence of self-reported STD was much lower in DC when compared with TC in both 2004 and 2005 (respectively, adjusted OR = 0.22 and 0.11, P <0.05, Table 3).
There is a trend of increasing prevalence of IDU behavior in TC (from 3.5% in 2003 to 10.9% in 2005, P <0.01) but not in DC (decreased from 9.5% to 5.6%, P = 0.18, Table 3). After adjusting for background factors, differences in prevalence in the 2 sites for the 3 surveys were not statistically significant (Table 3).
Health Services Utilization in DC Versus TC
At the baseline, 64.5% of the respondents had had an STD examination in the last 6 month in DC and the prevalence increased to 75.4% in 2005 (P <0.05, χ2 test for trend, Table 3) but no significant increase was observed in TC (31.2% and 37.5% in 2003 and 2005 respectively); adjusted OR comparing DC versus TC were significant for all 3 years (OR = 3.6–5.1, P <0.05). The prevalence of HIV antibody test in DC increased from 33.5% in 2003 to 75.4% in 2005 (P <0.01, χ2 trend test) but remained low in TC throughout the 3 years (8.9%, 7.8% and 3.6% respectively, P >0.05, χ2 trend test); adjusted OR comparing DC versus TC were 3.9, 62.4, and 84.5 respectively in the 3 years (P <0.05). Similarly, the prevalence of ever having received HIV preventive services (other than STD examination) increased from 64.5% in 2003 to 97.8% in 2005 in DC (P <0.01) but dropped from 26.2% in 2003 to 7.3% in 2005 in TC (P <0.001). The adjusted OR comparing DC and TC in 2004 and 2005 were also highly significant (Table 3).
Factors Predicting Condom Use With Sex Work Clients and Regular Sex Partners
It is seen that year of the survey, age group, education level, type of FSW, HIV-related knowledge level, belief that douching vagina after sex could prevent HIV, belief that external chemical contraceptives could prevent HIV, self-reported STD, possession of condom at the time of the interview, number of clients last week, IDU behavior, experience of HIV antibody testing and having received HIV-related preventive services were multivariately associated with either whether using condom with the last sex work client or consistent condom use with sex work clients in the last month (Table 4). Factors associated with condom use with regular sex partners are comparable to those associated with condom use with sex clients (Table 4).
Effective programs for HIV prevention are in dire need in China. However, most preventive programs have not been formally evaluated. A few evaluation studies have been reported using randomized control study design32,33 cohort or case-control design34,35 or descriptively7,36–38 These studies, however, measured efficacy rather than effectiveness and only evaluated programs at an individual rather than community level. No similar evaluation at a community level has been reported in China. The inclusion of a control community is also important.
It is seen that BSS data can be used to evaluate overall effectiveness of HIV prevention activities at a community level. The 2 places started out more or less the same in terms of HIV-related knowledge, condom use, and IDU behaviors, whereas significant improvements were observed in the intervention site (DC) when compared with the control site (TC)—apparent program effectiveness of the ECIR have therefore been observed in DC. This was further supported by the higher prevalence of self-reported STD in TC versus DC. From another source, the prevalence of syphilis in DC in 2003, 2004, and 2005 were respectively 8.3% (16/194), 6.3% (25/399), and 1.9% (3/161) (personal communication with the director of CDC in Dazhou). These prevalence data were therefore in agreement with the self-reported trend of STD of this study. Such data were, however, unavailable in TC.
Internationally, there were only a few reports using BSS data for similar evaluations. This may partially be attributed to the lack of expertise for data analyses at the local level in developing countries. Development of standardized user-friendly statistical and graphical programs for local BSS teams may therefore facilitate utilization of BSS data. This report may serve as a starting point.
Effective HIV prevention efforts are therefore most needed to dampen any further increases of HIV infection among FSWs in Dazhou. In 2003, 2004, and 2005 respectively, the results of the biologic surveillance for FSWs showed that the HIV prevalence were 1.0%, 4.0%, and 10.6% in DC (personal communication with the director of Dazhou CDC). Biologic surveillance was not conducted in TC. Therefore, despite an increasing trend of increasing condom use in DC, HIV prevalence among FSWs in DC was also increasing. This may be due to the trends of high prevalence of IDU behaviors in the sampled FSWs in DC and the trend of increasing HIV prevalence among IDUs in DC over time (52.9% in 2002 vs. 69.7% in 2005, personal communication with the director of Dazhou CDC). Furthermore, a time lag is expected before HIV prevalence stabilizes or declines, even if new effective prevention programs are to be provided, as the HIV-positive FSWs may have contracted HIV several years ago. Biologic surveillance of HIV alone is therefore inadequate. BSS for monitoring locales like DC and TC, where the likelihood of sharp increases in HIV prevalence exist, are therefore greatly warranted.
The results of the study show that it is possible to bring about substantial behavioral changes in risky sexual behaviors among FSWs within a timeframe of a few years. The effectiveness of the integrated programming and surveillance efforts serves as a model for designing HIV prevention programs in China. However, there are still concerns about the coverage and sustainability of effective prevention programs in China. More of such programs are envisaged as a result of international and national efforts. This study also reminds us that it may be problematic to extrapolate BSS results obtained from areas with intensive intervention programs (such as ECIR) to the majority of other areas which have not yet been covered by intensive intervention programs.
The results suggest that the studied FSWs served as an important bridge population for HIV transmission, as a noteworthy proportion of them were also IDUs. Moreover, these FSWs were less likely than other FSWs to be consistent condom users with both their sex work clients and their regular sex partners. Hence, the risk for HIV transmission from IDUs to sex work clients and regular sex partners of FSWs is alarming.18,19 Furthermore, inconsistent condom use with clients and with regular sex partners were significantly correlated with each other, thus increasing the chance of bridging HIV transmission between the sex work client population and the regular sex partner population via the IDU population.
Exposure to HIV-related preventive services in the past year was strongly associated with condom use with both sex work clients and regular sex partners, again suggesting that HIV preventive services have been effective in promoting condom use. Misconceptions (e.g., use of oral contraceptives is efficacious in HIV prevention) are, however, still quite common and were significantly associated with inconsistent condom use. Further, HIV-related knowledge was significantly associated with condom use with sex work clients. In the literature, the results have been mixed.39,40 Therefore, HIV preventive services in Sichuan should still promote HIV-related knowledge and rectify relevant misconceptions.
The possession of condoms at the time of the survey was significantly associated with condom use with sex work clients. There are worries that possession of condoms among FSWs may be used as evidence for prosecution as sex work is illegal in China. HIV prevention workers should therefore liaise with the public security officials to resolve this issue.
Condom use was very often associated with characteristics of respondents. Yet, fluctuations in the composition of the study populations (e.g., street workers) may occur across different surveys. Statistical adjustments of background characteristics are therefore required in interpreting data trends. Although certain responses may be subjected to reporting bias due to social desirability, the bias should have been reduced by anonymity of the survey. However, there is no reason to believe that the magnitude of such reporting biases is time-dependent or place-dependent. Selection bias may also exist even though mapping exercises have been conducted. The size of the biases in the intervention community should therefore be comparable to those of the controlled community, so that they are offset in relevant comparisons.
The one point of 2004 in TC (condom use with clients) which seemed to be lower than expected, and the data were cross-checked but no explanation could be offered. Though such could occur when cross-sectional surveys are to be compared, it reminds readers to interpret results with some caution. The remaining data were, however, quite consistent.
In conclusion, results obtained from these carefully planned and implemented BSS studies documented the effectiveness of the ECIR in promoting condom use among FSWs in a county in Dazhou and documented the level of HIV-risk behaviors. This report may serve as a model for applying BSS data to evaluate the overall program effectiveness in a community. These data can be used for both quality control for funding agencies as well as policy advocacy purposes. Insights have also been gained on the correlates of condom use. Integrating BSS with intervention programs and careful documentation of such programs in details will benefit HIV prevention work in different countries. These data are also important for modeling the HIV epidemic.41 BSS data have readily been collected in many countries but may often have been much underutilized.
1. Lu F, Wang N, Wu Z, et al. Estimating the number of people at risk for and living with HIV in China in 2005: Methods and results. Sex Transm Infect 2006;82 (Suppl III):iii87–iii91.
2. Wu FS. International non-governmental actors in HIV/AIDS prevention in China. Cell Res 2005; 15:919–922.
3. Community-level HIV intervention in 5 cities: Final outcome data from the CDC AIDS Community Demonstration Projects. Am J Public Health 1999; 89:336–345.
4. Jordan-Harder B, Maboko L, Mmbando D, et al. Thirteen years HIV-1 sentinel surveillance and indicators for behavioural change suggest impact of programme activities in south-west Tanzania. AIDS 2004; 18:287–294.
5. Laga M, Alary M, Nzila N, et al. Condom promotion, sexually transmitted diseases treatment, and declining incidence of HIV-1 infection in female Zairian sex workers. Lancet 1994; 344:246–248.
6. van Griensven GJ, Limanonda B, Ngaokeow S, et al. Evaluation of a targeted HIV prevention programme among female commercial sex workers in the south of Thailand. Sex Transm Infect 1998; 74: 54–58.
7. Wong ML, Chan RKW, Koh D. Promoting condoms for oral sex: Impact on pharyngeal gonorrhea among female brothel-based sex workers. Sex Transm Dis 2002; 29:311–318.
8. Wong ML, Chan R, Koh D. Long-term effects of condom promotion programmes for vaginal and oral sex on sexually transmitted infections among sex workers in Singapore. AIDS 2004; 18:1195–1199.
9. Amon J, Brown T, Hogle J, et al. Behavioral Surveillance Surveys BSS: Guidelines for Repeated Behavioral Surveys in Populations at Risk of HIV. Arlington, VA: Family Health International; 2000.
10. Family Health International and UNAIDS. Meeting the Behavioural Data Collection Needs of National HIV/AIDS and STD Programmes. A Joint IMPACT/FHI/UNAIDS Workshop: Report and Conclusions. Arlington, VA: Family Health International; 1998.
11. Centers for Disease Control and Prevention, United States. National HIV Behavioral Surveillance. Program in Brief. January 2004. Available at: http://www.cdc.gov/programs/hiv10.htm
. Accessed August 28, 2005.
12. 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.
13. Tung ND, Tuan NA, Hoang TV, et al. HIV/AIDS behavioral surveillance survey Vietnam 2000 (BSS Round 1 Results). USAID/FHI/Impact/ NASB; 2001.
14. Utomo B, Dharmaputra NG, Haryanto B, et al. Indonesia HIV/AIDS Behavioral Surveillance Survey: Results from the Cities of North Jakarta, Surabaya, and Manado. Jakarta: Center for Health Research; 1997.
15. Family Health International. HIV/AIDS Behavioral Surveillance Survey: Nigeria 2000 (Report and Data Sheets). Arlington, VA: Family Health International.
17. Qian ZH, Vermund SH, Wang N. Risk of HIV/AIDS in China: Subpopulations of special importance. Sex Transm Infect 2005; 81:442–447.
18. Choi SY, Cheung YW, Chen K. Gender and HIV risk behavior among intravenous drug users in Sichuan Province, China. Soc Sci Med 2006; 62:1672–1684.
19. Lau JT, Feng T, Lin X, et al. Needle sharing and sex-related risk behaviours among drug users in Shenzhen, a city in Guangdong, southern China. AIDS Care 2005; 17:166–181.
20. Liu H, Grusky O, Li X, et al. Drug users: A potentially important bridge population in the transmission of sexually transmitted diseases, including AIDS, in China. Sex Transm Dis 2006; 33:111–117.
21. UNAIDS/WHO. AIDS Epidemic Update: December 2005. Geneva: UNAIDS/WHO; 2005.
22. The UN Theme Group on HIV/AIDS in China. HIV/AIDS: China’s Titanic Peril—2001 Update of the AIDS Situation and Needs Assessment Report. Beijing: UNAIDS; 2002.
23. Lu L, Jia MH, Zhang XB, et al. Analysis for epidemic trend of acquired immunodeficiency syndrome in Yunnan Province of China. Chin J Prev Med 2004; 38:309-312 [in Chinese].
24. van den Hoek A, Yuliang F, Dukers NH, et al. High prevalence of syphilis and other sexually transmitted diseases among sex workers in China: Potential for fast spread of HIV. AIDS 2001; 15:753–759.
25. Chen XS, Yin YP, Liang GJ, et al. Sexually transmitted infections among female sex workers in Yunnan, China. AIDS Patient Care STD 2005; 19:853–860.
26. Joint China Ministry of Health and UN Theme Group on HIV/AIDS in China. China responds to AIDS: HIV/AIDS situation and needs assessment report. Bejing: Chinese Ministry of Health; 1997.
27. National Bureau of Statistics of China. China Statistical Yearbook 2005. Beijing: China Statistics Press; 2005.
29. China-UK HIV/AIDS Prevention and Care Project in Sichuan. Comprehensive surveillance report for HIV/AIDS, Sichuan 2003. Sichuan, China: UK HIV/AIDS Prevention and Care Project office; 2004.
31. China Center for Disease Control and Prevention. Guidelines for Comprehensive Surveillance of HIV/AIDS. China: CDC; 2002 [in Chinese].
32. Dolan KA, Shearer J, MacDonald M, et al. A randomized controlled trial of methadone maintenance treatment versus wait list control in an Australian prison system. Drug Alcohol Depend 2003; 72:59–65.
33. Sees KL, Delucchi KL, Masson C, et al. Methadone maintenance vs 180-day psychosocially enriched detoxification for treatment of opioid dependence: A randomized controlled trial. JAMA 2000; 283:1303–1310.
34. Kwiatkowski CF, Booth RE. Methadone maintenance as HIV risk reduction with street-recruited injecting drug users. J Acquir Immune Defic Syndr 2001; 26:483–489.
35. Moss AR, Vranizan K, Gorter R, et al. HIV seroconversion in intravenous drug users in San Francisco, 1985–1990. AIDS 1994; 8:223–231.
36. Abbott PJ, Moore B, Delaney H, et al. Retrospective analyses of additional services for methadone maintenance patients. J Subst Abuse Treat 1999; 17:129–137.
37. Celentano DD, Nelson KE, Lyles CM, et al. Decreasing incidence of HIV and sexually transmitted diseases in young Thai men: Evidence for success of the HIV/AIDS control and prevention program. AIDS 1998; 12:F29–F36.
38. Ghys PD, Diallo MO, Ettiegne-Traore V, et al. Increase in condom use and decline in HIV and sexually transmitted diseases among female sex workers in Abidjan, Cote d’Ivoire, 1991–1998. AIDS 2002; 16:251–258.
39. Fritz, Richard B. AIDS knowledge, self-esteem, perceived AIDS risk, and condom use among female commercial sex workers. J Appl Soc Psychol 1998; 28:888–911.
40. Wong ML, Lubek I, Dy BC, et al. Social and behavioural factors associated with condom use among direct sex workers in Siem Reap, Cambodia. Sex Transm Infect 2003; 79:163–165.
41. Brown T, Peerapatanapokin W. The Asian epidemic model: A process model for exploring HIV policy and programme alternatives in Asia. Sex Transm Infect 2004;80 (Suppl 1):i19–i24.