O'Farrell, Nigel MSc, MD, FRCP*†; Oula, Ratthiphone MSc, MD‡; Morison, Linda BSc†; Van, Cao Thi Bao MD, PhD§
INTERVENTIONS FOCUSING ON INCREASED condom use and improved sexually transmitted infection (STI) control in female sex workers (FSWs) are among the most cost-effective for HIV prevention.1 In Asia, FSWs have played a crucial role in the spread of HIV in Thailand and Cambodia and in both countries, programs that included comprehensive STI clinical services with free treatment focusing on brothel-based or direct FSWs have met with a considerable degree of success.2,3 By contrast, in Laos, sex work is clandestine and much less open than in neighboring countries. Commercial sex tends to take place away from public scrutiny and has been subject to police clampdowns in recent times.4 Few FSWs in Laos identify themselves as such and the term “service women” is regarded as a more acceptable occupational description.
In Laos, there are no designated STI clinics or clinical STI specialist clinicians. Service women with STIs that consult through the public sector tend to be seen by gynecologists, often with limited STI experience. For women presenting with vaginal discharge, including service women, syndromic management is used to diagnose cervical infections using a risk assessment score devised for use among the general population.
For many years, STIs have received little attention in Laos. Official reports to the Laos Ministry of Health identified 2624 STI cases throughout the whole country in 2001.5 However, STI prevalence surveys in Laos have detected very high prevalences in service women. In 2001, the HIV Sentinel Surveillance and Sexually Transmitted Infection Periodic Prevalence Survey (HSS-SPPS) detected respective prevalences of gonorrhea and chlamydia in Vientiane, Luang Prabang, and Savannakhet of 14% and 16%, and 13% and 35%, 32% and 23%,6 whereas even higher rates were found in Luang Prabang in 2003 where baseline prevalences of chlamydia (46%) and gonorrhea (34%) were detected in a presumptive treatment (PT) study among FSWs.7 Both studies used the Roche Amplicor method for detecting gonorrhea and chlamydia using urine specimens in the former and cervical in the latter.
Despite these high STI levels, HIV prevalence in Laos has been low even in high-risk groups. Among service women, the HSS-SPPS survey detected HIV prevalences of 1% in Vientiane and Savannakhet and no cases were found in Luang Prabang.6 Clearly, given these very high STI levels in service women, there is the potential for a significant HIV epidemic to develop.
To mitigate the risks of HIV in populations at high risk, the Community Action for Preventing HIV/AIDS project, Japanese Fund for Poverty Reduction (JFPR-9006), managed by the Asian Development Bank (ADB) was conceived to target selected border provinces in 3 Mekong countries: Vietnam, Cambodia, and Laos. The project comprised an extensive HIV prevention package involving information, education and communication, condom promotion, and improved STI control both in the general population and FSW focusing on mobile groups.8 Because of the difficulties in accessing service women in Laos, periodic presumptive treatment (PPT), also sometimes described as periodic asymptomatic treatment, was chosen as the main STI intervention strategy for this group in the 3 project border provinces of Oudomxai, Khammouane, and Savannakhet. PPT entails treating a group at high risk of infection at regular intervals regardless of symptoms based on the premise that many will have untreated infections. STI management is accomplished without the need for laboratory tests other than to perform surveillance as part of program evaluation.
Acceptance of Periodic Presumptive Treatment
To implement PPT in the provinces, the strategy had to be accepted by the Ministry of Health at the highest level for ethical approval. This was accomplished after an advocacy meeting with the Minister of Health. This meeting also involved the Director of Hygiene & Prevention Department, Education Department, Transport & Communication Department, police, military, 3 Provincial Governors, Lao Women's Union, Lao Trade Union, Director of the Centre of Laboratory and Epidemiology–Deputy Director of the National Ethical Committee, Lao Tourism Authority, Lao Youth Union, and the Directors of the 3 Provincial Health Departments. Further advocacy meetings were held locally in the 3 provinces (Oudomxai, Khammouane, and Savannakhet) and involved a similar wide spectrum of community groups as those that attended the national workshop.
At the province level, 2 working groups, one for technical issues and the second for monitoring and evaluation, were formed to cover both PPT and 100% condom use programs. Two teams were set up within each province, a technical team to focus on STI issues comprising doctors and nurses from the provincial hospital and a behavior change communication team with a wide range of collaborators including members of the Provincial Health Team, Lao Women's Union, police, Lao Tourist Board, and Lao Youth Union.
Training workshops lasting 1 week were then held in the 3 provinces. The basis of these was to use experience gained from a PT program in Luang Prabang.7 In addition, meetings were held with owners and managers of entertainment establishments to explain further about the intended benefits for the service women from the project.
Based on the assumptions of statistical significance at the 5% level and 90% power, a 40% baseline prevalence of having either gonorrhea and/or chlamydia (GC/CT) and an expected 50% reduction in the prevalence of the same, a minimum sample size of 180 was required.
During the project baseline survey in April 2002, mapping was undertaken of establishments where service women were likely to be found. These locations were mainly small drink shops, hotels, guest houses, restaurants, and nightclubs. This mapping was done again in mid-2004 before the start of the intervention. Service women were included if they were aged 15 years or more with access to clients. The study was explained and informed verbal consent was obtained from all subjects. All those invited participated in the study.
The numbers of service women recruited varied between the provinces. A “take all” approach was adopted. In Oudomxai, service women were recruited in Muang Xai, the main town of Oudomxai Province. In Khammouane, service women were identified in Thakhaek, the main provincial town. In Savannakhet, subjects were selected from those places that had not been used in the cluster sampling for the HSS-SPPS survey in September 2004 in Khanthabury, the main district in the center of Savannakhet town.9 The estimated numbers of sex workers in the 3 areas sampled in these three provinces were 100, 160, and 400, respectively.
Preceded by a light snack, all subjects were given 1 g of azithromycin to cover possible cervical infections with GC/CT and were observed taking it. Presumptive treatment for syphilis was not considered because the prevalence in service women is very low (0.2%).6
Service women were interviewed in a face-to-face discussion with a trained interviewer in a confidential area. A standard questionnaire was used to ask about sociodemographic details, sexual activity, if they had sold sex, condom use with both clients and other partners, and current STI symptoms.
The first round of azithromycin was administered in August 2004 in both Khammouane and Oudomxai and in October 2004 in Savannakhet. All the service women received counseling and health education about safer sex and were provided with free condoms. Subjects with STI symptoms were advised to attend local clinics if still symptomatic after 1 week.
After the interview, urine samples were collected from subjects who had been advised not to pass urine in the previous 2 hours. A first-catch urine (the first part of the stream) was collected into clean polypropylene containers (Corning) without preservatives. After being appropriately sealed and labeled, specimens were transported to the main local laboratory in an icebox and stored at 4 to 8°C initially for a maximum of 72 hours, then at the Provincial Centre for the control of AIDS at −20°C, and subsequently to the Central Laboratory in Vientiane at −70 °C. Specimens were then packed in dry ice and sent in 2 batches by plane to Ho Ch Minh City, Vietnam. On arrival, specimens were immediately transported to the Pasteur Institute, Ho Chi Minh City, and tested for gonorrhea and chlamydia by polymerase chain reaction (PCR) (Amplicor; Roche, Branchburg, NJ).
Gonorrhea-positive specimens were confirmed with an in-house method using specific primers to amplify a 273-bp fragment from the plasmid-encoded cppB region that is part of the 390-bp sequence on the same plasmid.10,11 Detection was by hybridization using a digoxigenin probe. Inhibited samples underwent DNA purification using the QIAamp DNA mini kit (Qiagen) and in-house PCR testing.
Data from the interviews and the results of laboratory tests were entered into EpiInfo Version 6.04 and analyzed. STATA version 8 was used to calculate confidence intervals. Prevalence rates of GC/CT were compared using the chi-squared test for comparing proportions.
The initial plan was to give PPT at monthly intervals on day 1 and subsequently on days 30, 60, and 90 (Table 1) with urine samples to be collected on days 1 and 90. This was accomplished in Khammouane but not in Savannakhet and Oudomxai where the second- and third-round courses of PPT, respectively, were not given as a result of unforeseen staff shortages.
Urine samples were collected in the 3 provinces on days 1 and 90. The numbers of urine samples did not match the numbers of questionnaires in Oudomxai on day 1 because 17 service women there were Chinese and did not understand Lao sufficiently to answer the questionnaire. Otherwise, except for 4 service women unable to pass urine on both day 1 and day 90 in Khammouane and 3 unable to wait to answer the questionnaire in Savannakhet, the numbers of urine samples matched the numbers of questionnaires completed. All participants accepted azithromycin at each round.
The numbers of service women recruited are shown in Table 2. Four hundred forty-two women were interviewed at baseline and 419 on day 90. The numbers of service women seen on day 90 that had not had any previous rounds of PPT was 125 of 200 (62.5%) in Savannakhet and 13 of 65 (20%) in Oudomxai, but was not recorded in Khammouane.
Sociodemographic and Sexual Behavior Characteristics
The majority of service women in all three provinces were aged less than 25 with the overall greatest number (45.8%) in the 15- to 19-year age group. Service women tended to be younger in Oudomxai and Savannakhet than Khammouane (Table 2).
The location of service women varied considerably both within and between the provinces. In Khammouane, on day 1, 22.2% were enrolled from small drink shops and 3.7% from hotels compared with 45.8% and 27.7%, respectively, at these places on day 90.
In all 3 provinces, significant proportions gave their home province as somewhere other than the province where they were currently staying (Table 2). This was greatest in Oudomxai and Khammouane where approximately half of the service women interviewed were from other provinces originally. The pattern of mobility varied across the 3 provinces, but there was significant movement from Luang Prabang to Oudomxai, from Vientiane Capital and Sekong to Savannakhet, and from Boikamxay and Vientiane to Khammouane.
All subjects interviewed in Savannakhet admitted having had sexual intercourse. In Oudomxai, 7.7% on day 90 and 11.1% in Khammouane on day 1 denied previous sexual intercourse. Overall, 122 of 861 (14.2%) of those interviewed in the 3 provinces over days 1 and 90 reported they had not exchanged sex for money in the past year.
Reported 100% (consistent) condom use with regular partners in the past month varied between the provinces from 41.8% in Oudomxai on day 1 to 90.8% in Savannakhet on day 90. Reported condom use with clients at last sex was greater in Savannakhet than in the other 2 provinces. Consistent condom use with clients in the last month was reported more frequently in Savannakhet at approximately 75% on both day 1 and day 90, although this variable increased from 62% in Oudomxai and 58.6% in Khammouane on day 1 to 72.3% and 65.2%, respectively, on day 90.
Sexually Transmitted Infection Symptoms and Infections
A total of 42.3% of the service women reported current STI symptoms on day 1 and 38.0% on day 90. Abnormal vaginal discharge was the most frequent symptom and was reported by 22.6% on day 1 and 27.4% on day 90.
The prevalences of gonorrhea, chlamydia, and GC/CT at day 1 and day 90 are shown in Table 3. Twenty-eight inhibited samples were found on the first round of testing. The highest baseline levels were found in Oudomxai and Khammouane. The largest drop in the prevalence of GC/CT between day 1 and day 90 was seen in Oudomxai (0.35, P <0.001) followed by Khammouane (0.55, P = 0.002) and Savannakhet (0.75, P = 0.15) with an overall reduction of 0.55 (P <0.001).
Some of the service women found to have gonorrhea, chlamydia, and GC/CT reported never having had sex (Table 4). Nineteen of 223 (8.5%) women interviewed and found to have GC/CT on either day 1 or 90 claimed not to have had sex. Overall, 188 of 223 (84.3%) of those with GC/CT reported ever selling sex.
In all provinces, a significant proportion of GC/CT infections were identified in those reporting consistent condom use with clients. In both Khammouane and Savannakhet, this proportion increased from 21.9% and 28.6% on day 1 to 54.5% and 50%, respectively.
This study has demonstrated a high prevalence of both gonorrhea and chlamydia in service women in 3 border provinces in Laos. After the implementation of a short period of PPT with azithromycin, the prevalences were lowered significantly. Higher STI baseline levels were recorded in Oudomxai and Khammouane compared with Savannakhet and might reflect both longer established STI and 100% condom use programs in the latter province. In Khammouane, the ADB project was the first to address HIV and STI-related issues in the province, whereas in Oudomxai, a previous STI project had finished some time ago and STI activities had not been sustained. Also, service women in Oudomxai were younger than those in the other 2 provinces and would not have been exposed directly to the earlier project.
The overall prevalence of cervical infections was reduced by 45% from 32.4% on day 90 to 18.0% with reductions of 71% in Oudomxai, 45% in Khammouane, and 25% in Savannakhet. The variation in reduction did not seem to be explained either by obvious differences in the distributions of the characteristics of service women at day 1 and day 90 or the different number of times PPT was given in the 3 provinces before day 90 (3 times in Khammouane and twice in the other 2 provinces). It might rather reflect the more stable pattern of service women in Oudomxai where 20% of subjects seen on day 90 were new attenders compared with 62.5% in Savannakhet. Mobility was also thought to be significant in Khammouane and is likely to become an increasingly important factor in STI/HIV transmission now that the Nam Theun Dam construction project in the east of the province has started. Clearly, these mobile service women may not be aware of how to access STI care when they arrive in new surroundings and are particularly vulnerable. Mobility was also a factor within provinces and could account for some of the variations in workplace between rounds in each province. Establishments were liable to open and close on a regular basis depending on local governmental and police restrictions.
The numbers reporting current STI symptoms overall, particularly abnormal vaginal discharge, was quite high and similar on day 1 (22.6%) and day 90 (27.4%) despite the lower cervical STI prevalence at day 90. This supports the contention that reported abnormal vaginal discharge is a poor predictor of cervical infections in FSWs.12,13 Whether this reflects other reproductive tract infections or psychosocial factors, as identified in a community-based survey in India, is unknown.14
A significant proportion of women were infected with GC/CT in this study despite reporting consistent condom use with clients. Although this might reflect inaccurate reporting or false-positive STI tests, it could also mean that service women were getting reinfected from regular noncommercial partners or that STI rates in clients are very high. Clearly, male contacts of service women are an important group to target in Laos. However, the largest group of clients of service women is government workers or businessmen, groups that are difficult to reach in this setting.6
In Laos, like in most other countries in Asia, men tend to obtain STI drugs directly from pharmacies. However, tablets usually dispensed for STIs over the counter in Laos such as chloramphenicol, thiamphenicol, and kanamycin, none of which are recommended by World Health Organization STI Guidelines, have limited efficacy for STIs.15 Furthermore, absorption of kanamycin is minimal when given orally. Before the PPT project, neither azithromycin nor cefixime, the latter being the first-line oral treatment for gonorrhea in many countries both in Asia and the West, were available in Laos. Subsequently, they were both licensed and other cheaper versions of these 2 drugs have since become available. Although it may be some time before azithromycin and cefixime are stocked routinely in pharmacies in Laos, this will be an essential step toward reducing these high STI levels. Combination packages of azithromycin and cefixime specifically for urethral discharge/dysuria syndrome would seem an attractive option and negotiations are currently underway with statutory medical agencies for legal approval for their use over the counter. The finding of high levels of chlamydia in male truck drivers—up to 16% in Luang Namtha province—emphasizes the need for urgent action.9
Some of the answers to the questionnaires here must remain open to question. For example, 8.5% of service women with GC/CT overall denied ever having had sexual intercourse. Some service women reported selling sex but denied sexual intercourse, and it may be that nonpenetrative sex was sold. Similar findings were reported in the HSS-SPPS survey in which 5.5% of service women had chlamydia among those reporting never having had sex.6 In Laos, like in Africa, it would seem that the unreliability of self-reported sexual behavior16 reflects difficulties in defining sex work in general, hence our decision to enroll all service women in our study.
Patterns of sexual mixing in service women would appear to be changing rapidly in Laos as demonstrated by the increase in numbers of clients of service women recently. In 2001, the HSS-SPPS survey found that the mean numbers of partners per year in service women was 14.6 In the ADB exit survey, the mean numbers of sexual partners per week were 1.8 in Oudomxai, 1.6 in Khammouane, and 1.5 in Savannakhet.17 In Luang Prabang in 2003, the mean numbers of partners per week was 2.7
This study did not have a control group and it could be that there were seasonal differences, secular trends, variations in condom availability, or other confounding factors that accounted for the different GC/CT levels rather than PPT. However, seasonal variation is more likely to account for changes in gonorrhea,18 whereas the relative reduction in both gonorrhea and chlamydia was similar. One hundred percent condom use programs had been introduced earlier in the 3 provinces, and the project also supported both the distribution of free condoms for STI clients attending public services and an extensive condom social marketing program. Although health education was given to service women as part of the 100% condom use program, similar health educational material was also used for the PPT programs.
Studies of PPT for cervical infections in Asia are limited, although the strategy has been used in some countries without formal evaluation. In Fuzhou, China, chlamydia prevalence decreased from 27.5% to 5.9% with monthly 1 g azithromycin over 3 months.19 PPT with injections of benzathine penicillin twice monthly has been used for syphilis in FSWs in Indonesia, although prevalence was not reduced.20 In Africa, PPT with azithromycin for FSWs has been undertaken but with varying degrees of success. In Free State, South Africa, in a mining community, 1 g azithromycin was given monthly to FSWs and reduced the prevalence of GC/CT from 24.9% to 5.7% in those with 4 visits.18 In a South African mining area in Carletonville, gonorrhea and chlamydia rates decreased significantly among FSW attending a PPT program at least 4 times.21 In Nairobi, Kenya, monthly azithromycin reduced the incidence of both gonorrhea and chlamydia significantly.22 In Accra, Ghana, and Cotonou, Benin, the impact of PPT in FSWs in a randomized, controlled trial of 1 g azithromycin on day 1, followed by ciprofloxacin on day 30 and day 60, a cycle repeated 3 times was limited, probably because other intensive interventions had been implemented over the previous few years.23 A community mass treatment trial with azithromycin and ciprofloxacin at 9-month intervals for 3 rounds at the population level in Rakai, Uganda, failed to reduce gonorrhea and chlamydia.24
A simpler related STI control strategy to PPT is PT that involves a single dose of therapy to an at-risk population. However, initial benefits may not be sustained.25,26 Single-dose 1 g azithromycin was tried in Laos in service women in Luang Prabang, but despite an intensive lead-in program, high prevalences of both gonorrhea and chlamydia were still apparent 3 months after treatment was given.7 In Angeles, Philippines, a similar PT strategy with improved access to STI screening had a modest effect with levels of GC/CT of 22% to 52% at baseline decreasing to 23% to 34% after 6 months.27
Although there is concern about emerging gonorrhea resistance to azithromycin, the 1-g dose was also used in the South African and Nairobi studies where treatment failure was not thought to be an issue.18,21,22 Resistance to azithromycin has not been documented in the region and although World Health Organization guidelines recommended 2 g to cover gonorrhea,28 the real possibility of gastrointestinal side effects led us to use the 1-g dose. The fact that the reduction in gonorrhea was greater than that of chlamydia suggests that treatment failure with the 1-g dose was unlikely. However, antibiotic resistance patterns to gonorrhea should be monitored closely in the future.
The overall prevalence of GC/CT at 18.0% at day 90 in this study was still high and points to the need for intensive interventions incorporated into the healthcare infrastructure that will enable mobile and new sex workers to receive STI treatment on a sustained basis rather than through a short-term intervention such as the one here. The dramatic reduction in the combined cost of 1 g azithromycin and 400 mg cefixime from $15 to $3 in the last 2 years in Laos now means that monthly or bimonthly PPT is a cheap and sustainable option for STI control in this group. Clearly, with the high STI levels here, syndromic management for service women with vaginal discharge must be regarded as having limited value. Furthermore, relying on Gram stains to predict GC/CT is also unlikely to reduce STI levels to a satisfactory level. Some authorities recommend that PPT be used as early as possible in the development of an STI epidemic but also point out that ethical and political concern may preclude instituting such a policy.29 These concerns have now been addressed in Laos, and the way forward would appear to be to have new service women-friendly clinics or outreach posts where PPT could be given to new service women in a setting where safer sex counseling could be delivered in a more appropriate setting than public gynecology clinics. In the other JFPR-9006 project countries, Cambodia and Vietnam, mobility in FSWs is common,30,31 and PPT would seem to be an attractive STI control strategy to be adopted at the regional level. Worryingly, a recent increase in the prevalence of HIV has been identified in service women in Savannakhet on the Thai border and Bokeo province, bordering Myanmar and Thailand.9 Investment in STI control in service women with a national PPT program may be one important option in minimizing the spread of HIV in Laos while HIV levels remain low.
1. Moses S, Plummer FA, Ngugi EN, et al. Controlling HIV in Africa: Effectiveness and cost of an intervention in a high-frequency STD transmitter core group. AIDS 1991; 5:407–411.
2. Rojanapithayakorn W, Hananberg R. The 100% condom program in Thailand. AIDS 1996; 10:1–7.
3. Sano P, Sopheap S, Sun LP, et al. An evaluation of sexually transmitted infection case management in health facilities in 4 border provinces of Cambodia. Sex Transm Dis 2004; 31:713–718.
4. Lyttleton C. Any port in a storm: coming to terms with HIV in Lao PDR. Culture Health & Sexuality 1999; 2:115–130.
5. Ministry of Health. National Statistics. 2001.
7. NCCAB/FHI. Luang Prabang STI intervention, monitoring and evaluation findings. Presented at Dissemination meeting, Luang Prabang, September 23, 2004.
9. Second HIV Surveillance Survey and Sexually Transmitted Infection Periodic Prevalence Survey, Laos People's Democratic Republic 2004. Draft Report FHI. 2005.
10. Ho BS, Feng WG, Wong BK, Egglestone SI. Polymerase chain reaction for the detection of Neisseria gonorrhoeae
in clinical samples. J Clin Pathol 1992; 45:439–442.
11. Tabrizi SN, Paterson B, Fairly CK, et al. A self-administered technique for the detection of sexually transmitted diseases in remote communities. J Infect Dis 1997; 176:289–292.
12. Wi T, Mesola V, Manalastas R, et al. Syndromic approach to detection of gonococcal and chlamydia infections among female sex workers in two Philippine cities. Sex Trans Infect 1998; 74(suppl 1): S118–122.
13. Diallo MO, Ghys PD, Vuylstke B, et al. Evaluation of simple diagnostic algorithms for Neisseria gonorrhoeae
and Chlamydia trachomatis
cervical infections in female sex workers in Abidjan, Cote d'Ivoire. Sex Transm Infect 1998; 74(suppl 1): S106–111.
14. Patel V, Pednekar S, Weiss H, et al. Why do women complain of vaginal discharge? A population survey of infectious and psychosocial risk factors in a South Asian community. Int J Epidemiol 2005; 34:853–862.
15. O'Farrell N. Trip Report. STIs in Laos. Asian Development Bank, July 2002.
16. Morison L, Buve A, Zekeng L, et al. Commercial sex and the spread of HIV in four cities in sub-Saharan Africa. AIDS 2001; 15(suppl 4): S61–S69.
17. Oula R. Presented at Inter-country meeting, JFPR-9006, March 14–17, 2005, Siem Reap, Cambodia.
18. Steen R, Vuylsteke B, DeCoito T, et al. Evidence of declining STI prevalence in a South African mining community following a core group intervention. Sex Transm Dis 2000; 27:1–8.
19. Ping Y, Xiaohong S, Yueping Y, et al. Evaluation of mass treatment for STI with oral azithromycin among unlicensed prostitutes in China. 13th International Union against Sexually Transmitted Infections Asia Pacific Conference, July 6–9 2004, Chiang Mai, Thailand; Abstract O22.
20. Joesoef MR, Valleroy LA, Kuntjoro TM, et al. Risk profile of female sex workers who participate in a routine penicillin programme in Subrabaya, Indonesia. Int J STD AIDS 1998; 12:756–760.
21. Steen R, Olivier E, Mzaidume Z, et al. Periodic presumptive treatment as a component of a comprehensive HIV/STD prevention programme in a South African mining community. ISSTDR, Berlin, June 2001.
22. Kaul R, Kimani J, Nagelkerke NJ, et al. Monthly antibiotic chemoprophylaxis and incidence of sexually transmitted infections and HIV-1 infection in Kenyan sex workers. JAMA 2004; 291:2555–2562.
23. Labbé A-C, Dzokoto A, Khonde N, et al. A randomised placebo-controlled trial of routine monthly antibiotics against gonococcal and chlamydial infections among female sex workers in Ghana and Benin: Intention-to-treat analysis. International Society for Sexually Transmitted Disease Research, Ottawa, July 2003; Abstract 0524.
24. Wawer MJ, Sewankambo NK, Serwadda D, et al. Control of sexually transmitted diseases for AIDS prevention in Uganda: A randomised community trial. Lancet 1999; 353:525–535.
25. Holmes KK, Johnson DW, Kvale PA, et al. Impact of a gonorrhoea control program, including selective mass treatment, in female sex workers. J Infect Dis 1996; 174(suppl 2):S230–239.
26. Rekart ML, Patrick DM, Chakraborty B, et al. Targeted mass treatment for syphilis with oral azithromycin. Lancet 2003; 361:313–314.
28. Guidelines for the Management of Sexually Transmitted Infections. Geneva: World Health Organization, 2001.
29. Wasserheit JN, Aral SO. The dynamic topology of sexually transmitted disease epidemics: Implications for prevention strategies. J Infect Dis 1996; 174(suppl 2):S201–213.
30. Sopheap H, Gorbach PM, Gloyd S, Leng HB. Rural sex work in Cambodia: Work characteristics, risk behaviours, HIV and syphilis. Sex Transm Infect 2003; 79:e2. Available at: www.stijournal.com/cgi/content/full/79/4/e2
31. Tuan NA, Fylkesnes K, Thang BD, et al. HIV infection patterns and risk behaviours in different population groups and provinces in Vietnam: Close links between drug injection related and heterosexual epidemics. XV International AIDS Conference, Bangkok, 11–16 July 11–16, 2004; Abstract MoPeC3449.