Article Text

Original article
HIV and sexually transmitted infection-related risks among female sex workers in Mongolia: second-generation surveillance survey, 2011–2012
  1. Sergelen Munkhbaatar1,
  2. Bulbul Aumakhan2,
  3. Baigalmaa Jantsansengee3,
  4. Iliza Azyei1,
  5. Zolzaya Sanjaajamts2,
  6. Jugderjav Badrakh2,
  7. Tugsdelger Sovd4,
  8. Susanne Stromdahl5,
  9. Stefan Baral5,
  10. Oyunbileg Amindavaa1
  1. 1Global Fund Supported Projects on HIV/AIDS and TB, Ministry of Health, Ulaanbaatar, Mongolia
  2. 2HIV/AIDS and STI Surveillance and Research Department, National Center for Communicable Diseases, Ministry of Health, Ulaanbaatar, Mongolia
  3. 3Mongolian Field Epidemiology Training Program, Ministry of Health, Ulaanbaatar, Mongolia
  4. 4Monitoring, Evaluation and Internal Auditing Department, Ministry of Health, Ulaanbaatar, Mongolia
  5. 5Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland, USA
  1. Correspondence to Dr Bulbul Aumakhan, HIV/AIDS and STI Surveillance and Research Department, National Center for Communicable Diseases, Ministry of Health, Ulaanbaatar 210648, Mongolia; an_bulbul{at}


Background Surveillance studies among female sex workers (FSWs) in Mongolia have found no HIV but high rates of syphilis, ranging from 10.7% in 2002 to 20.8% in 2007.

Objectives To determine the current prevalence of HIV and syphilis among FSWs, and to identify syphilis risk factors.

Methods 761 FSWs were recruited by time–location sampling between 2 January and 29 March 2012 in Ulaanbaatar city, Darkhan-Uul, Orkhon, Khuvsgul and Dornod provinces in Mongolia. Participants were administered an anonymous structured survey. Blood samples were tested for HIV and syphilis. Bivariate and multivariate logistic regression analyses were performed to identify factors associated with syphilis.

Results The mean age of participants was 31 (median 30, interquartile range 24–38). HIV knowledge was modest: 41.3% correctly answered all questions. Consistent condom use with clients was reported by 49.1% of participants and exposure to HIV prevention programmes by 50.1%. Although no cases of HIV were found, syphilis prevalence was −27.8% (95% CI 24.3% to 31.7%). In multivariate analyses, variables associated with syphilis were younger age (age >30 years, OR=0.96, 95% CI 0.92 to 0.99) and occurrence of genital ulcer (OR=2.24, 95% CI 1.17 to 4.28).

Conclusions A syphilis epidemic continues to grow among FSWs in Mongolia. These women are at high risk of HIV transmission if introduced into their sexual networks. With the increase in migration of mining workers in Mongolia, introduction of HIV may be imminent. Efforts to intensify treatment and prevention programmes among FSWs are needed.


Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


More than three decades since HIV/AIDS emergence, Mongolia remains a country with a very low number of reported HIV cases. By the end of 2013, 150 HIV cases had been cumulatively reported in the country (HIV surveillance data, the National Center for Communicable Diseases) with an estimated prevalence among the general population of <0.1%.1 However, for the past two decades, in the wake of the fall of communist regime and collapse of Soviet-built centralised healthcare system, Mongolia has been dealing with consistently rising and high rates of sexually transmitted infections (STIs).2 ,3 STIs were the second most common infectious disease, contributing 33.5% of all infectious diseases registered in the country in 2012.4 Syphilis, gonorrhoea and trichomoniasis are the most commonly reported STIs, and the observed rates in 2012, based on case notifications, were 178, 192 and 150 per 100 000 population, respectively.2 Increase in the notification rate of STIs was particularly noticeable for syphilis. Syphilis re-emerged in the 1990s, increasing from 18/100 000 during 1983–1993 to 32/100 000 in 1995.5 The syphilis notification rate reached 71/100 000 in 2001 and further doubled to 152/100 000 in 2011.6 There are geographical differences in the distribution of reported syphilis cases. Specifically, in 2012, syphilis notification rates were highest in Ulaanbaatar city (240/100 000), Dornod (359/100 000) and Darkhan-Uul (211/100 000) provinces, which are major socioeconomic areas with densely populated urban centres, and lowest in Bayan-Ulgii province (21/100 000), which is the westernmost province in Mongolia with inhabitants comprising mostly ethnic Muslim Kazakhs.4 It should be noted, however, that the accuracy of STI surveillance data is not known and the true STI disease burden remains to be assessed.

Societal norms towards non-commercial heterosexual activity are fairly relaxed.7 Between 40% of young people age 15–24 and 61% of male STI clinic attendees participating in the 2007 survey reported having multiple sexual partners in the past 12 months.8 An increasing trend in reported STIs, together with fairly relaxed societal norms towards heterosexual behaviour, have been considered to indicate risk patterns that could facilitate the spread of HIV.7

Female sex workers (FSWs) have been most affected by STIs in Mongolia and are at increased risk of HIV.9 Sex work is illegal in Mongolia and concentrated in major urban centres and border towns. Anecdotal evidence suggests the existence of different types of sex workers, including those engaged in commercial sex activities abroad, escorts working in upscale hotels, massage parlours and saunas, sex workers operating in areas with big mining and construction projects and sex workers who solicit clients on the streets. The extent and nature of these different levels of sex workers and their sexual networks is not known. Programmes targeting sex workers are mainly supported by international donor funds and delivered by non-governmental, community based organisations (NGOs) and drop-in centres in Ulaanbaatar city and a few provinces. These programmes use the existing STI facility to provide comprehensive HIV and STI prevention services to sex workers. These include HIV/STI voluntary testing and counselling, social support, condoms, outreach and education.10 STI treatment and care in these centres and public clinics is provided free of charge.

Recognising the increased vulnerability of FSWs to STIs and HIV, and their potential role as a bridge to the general population, periodic sero-behavioural second-generation surveillance surveys (SGSS) have been conducted since 2002. The six rounds of SGSS conducted among FSWs found no HIV but a high syphilis rate, ranging from a low of 10.7% in 2002 to a high of 20.8% in 2007.5 In this report, we describe the 7th SGSS conducted among 761 FSWs in Ulaanbaatar city, Darkhan-Uul, Orkhon, Khuvsgul and Dornod provinces between 2 January and 29 March 2012. The goal of the survey was to determine the current prevalence of HIV and syphilis among FSWs, and to identify syphilis risk factors.


The survey procedures were approved by the scientific committee of the National Center for Communicable Disease (NCCD) and approved by the ethical committee of the Ministry of Health. Written consent was obtained from all participants before enrolment into the survey.

Survey design, sites and population

FSWs were recruited using time–location sampling (TLS) from Ulaanbaatar city, Darkhan-Uul, Orkhon, Khuvsgul and Dornod provinces. A review of the TLS method is available elsewhere.12 ,13 These provinces were selected because they are socioeconomically more developed, with densely populated urban centres, and have NGOs working with FSWs. A sex worker was defined as a female aged between 15 and 49 years who is present at a selected venue–day–time (VDT) cluster and reported receiving money or gift for sex services in the past 12 months.

Sample size

Using an OpenEpi14 tool and assuming that 2% of women aged between 15 and 49 years are sex workers,15 a design effect of 2, response rate for blood collection among behavioural survey respondents of 90%, and 17.8% syphilis rate observed among FSWs in Ulaanbaatar during the 2009 SGSS16 as the expected frequency of outcome factor, the target sample size for 95% CI was estimated to be 361 for Ulaanbaatar city. The estimated numbers of eligible of FSWs in the provinces were low and hence, the sample size for each province was set at 100 bringing the total sample size to 761.


A list of venues (streets, bars, restaurants, hotels, saunas, etc) where FSWs congregate was compiled with input from government officers, representatives from NGOs, social workers of drop-in centres and outreach FSWs. Monthly sampling calendars were created through random selection of venues in the list. Once a venue or associated VDT cluster was selected, the survey team visited the venue and for the first 30 min counted community members. The count then was multiplied by eight to arrive at the estimate of the number of FSWs who might attend the venue during 4 h of sampling. Study subjects were approached by field interviewers and recruited systematically.

Sampling was conducted on 307 occasions and 715 FSWs enrolled into the survey. To reach the target sample size and to facilitate recruitment of sex workers from venues that were hard to access (eg, saunas and massage parlours) special occasions for sampling were organised with the help of NGOs, during which 46 additional participants were recruited.

Data collection

Interview and biological specimen collection followed the recruitment. The interview lasted for about 30 min and was conducted at the location. The questionnaire covered basic demographics such as age, education level, marital status, sexual behaviour, knowledge of HIV transmission risks, HIV testing history and coverage by HIV prevention services. Sexual behaviour questions elucidated number and occupation of clients, income from sex services, condom use, substance use history and self-reported history of STIs.

Laboratory procedures

Blood samples were delivered to central laboratory at the NCCD and tested for HIV using rapid tests (SD HIV 1/2 Rapid 3.0, Standard Diagnostics Inc, Korea). Reactive specimens were confirmed using ELISA and western blot. Syphilis testing was done using the rapid plasma reagin test (RPR test kit, Newmarket Laboratories Ltd, Kentford, Newmarket, CB8 7PN, UK) and reactive specimens were confirmed with a Treponema pallidum haemagglutination assay (TPHA test kit, New Market Laboratories). A specimen was considered positive for syphilis if both TPHA and RPR tests were reactive, with RPR titre ≥1:2.

STI case management

Survey participants received counselling before and after testing. Participants with laboratory evidence of STI were managed at NCCD or respective provincial STI clinics and health centres free of charge using national STI diagnosis and treatment guidelines.

Data analysis

All questionnaires were double entered to minimise errors. Missing data comprised <10% for all characteristics (except the question on whether mosquito bites transmit HIV to which 15% responded “don't know”) and were not included in the analyses. Continuous variables were categorised into groups (eg, age, number of partners) based on the observed distribution or a priori defined categories of interest.

For weighted analysis a simplified approach based on enumeration of each period of sampling was used to produce probability weights.13 FSWs recruited by NGO-organised sampling were not recruited at randomly selected VDTs and thus, were assigned a weight of 1.0. To account for potential clustering by venues, place of residence (city or province) was designated as a cluster.

Bivariate associations were evaluated using χ2 or Fisher's exact test for categorical variables and t test for continuous variables. Characteristics associated with syphilis in a bivariate analysis at p<0.20 were entered in a multivariate model. Variables of theoretical interest (condom use at last sex and in the past 12 months, variable indicating correct answer status to all HIV transmission questions, etc) were entered into multivariate regression regardless of the p value. An age spline term at a cut-off point of 30 years was introduced into a regression model to account for non-linear association between age and syphilis variables. Weighted ORs were obtained using survey logistic regression method weighted on the syphilis variable.

Both crude and weighted estimates with 95% CI for weighted estimates are presented. Unweighted p values are provided except where there are significant differences at the p<0.05 level, in which case, a weighted p value (Pw) is also shown. All analyses were conducted using STATA statistical software (STATA/IC V.12.1, STATA Corporation, College Station, Texas, USA).


Respondent characteristics

The mean age of participants was 31 years (median 30, interquartile range (IQR) 24–38 table 1). Over one-third (36.6%) had only primary and/or less than complete secondary level of education. By marital status, 29.1% (weighted 30.7%, 95% CI 27.0% to 34.7%) were single and had no regular partner and 37.6% (weighted 37.6%, 95% CI 33.7% to 41.6%) were separated, widowed or divorced.

Table 1

Selected characteristics of female sex workers recruited via time location sampling in Mongolia (n=761)

Mean and median age at sexual debut was 18 (range 8–27). Mean number of sexual partners in the past week was four (median 3, range 0–31). Mean number of clients on the last day participant provided sex services was 2 (median 1, range 0–6). Condom use at last sex was highest with a client (81.2%, weighted 80.5%, 95% CI 77.1% to 83.4%) and lowest with a regular sex partner (19.4%, weighted 21.5%, 95% CI 15.6% to 28.9%). Just about half (49.1%, weighted 48.7%, 95% CI 44.7% to 52.8%) used a condom at all times with a client in the past 12 months. By client occupation, blue collar workers (57.4%, weighted 56.5%, 95% CI 52.4% to 60.4%) were the most common clients of participants in this survey. Occurrence of genital discharge in the past 12 months was reported by 36.1% (weighted 38.5%, 95% CI 34.6% to 42.5%) and genital ulcers by 7% (weighted 7.5%, 95% CI 5.5% to 10.1%) of participants.

Alcohol use was common with only 20.8% reporting rarely or never drinking (data not shown). Ever drug use was reported by 10 out of 758 participants (1.3%, weighted 1.1%, 95% CI 0.6% to 2.0%). Only one woman reported injecting drugs in the past 12 months.

The proportion of participants who responded correctly to individual questions testing knowledge of HIV transmission ranged from a low of 52.6% to a high of 87.9% (data not shown), while the proportion of those who responded correctly to all questions was 41.3% (weighted 42.9%, 95% CI 38.9% to 46.9%).

About 60% (59.3%, weighted 59.2%, 95% CI 55.0% to 63.2%) of participants reported being tested for HIV in the past 12 months and 93.5% received test results. About half of FSWs (50.3%, weighted 50.1%, 95% CI 46.1% to 54.2%) reported participating in HIV prevention programmes in the same period.

HIV and syphilis prevalence

No HIV infection was found among survey participants. Overall syphilis prevalence was 27.5% (weighted 27.8%, 95% CI 24.3% to 31.7%; table 2). Significant differences were found in the syphilis rate by survey sites. The highest prevalence of 36% (weighted 41.3%, 95% CI 30.5% to 52.9%) was found in Darkhan-Uul province followed by 30.5% in Ulaanbaatar city (weighted 31.3%, 95% CI 25.3% to 38.0%) and 31% in Dornod province (weighted 29.5%, 95% CI 20.0% to 41.2%). The prevalence values for Orkhon and Khuvsgul provinces were comparable: 21% (weighted 21.9%, 95% CI 14.6% to 31.6%) and 22% (weighted 18.3%, 95% CI 11.8% to 27.4%), respectively. Three women among 46 participants (6.5%, 95% CI 2.1% to 19.1%) recruited at NGO-organised events tested positive for syphilis. It should be noted, that TPHA was positive for 40.1% and RPR for 42.2% of participants, indicating high lifetime syphilis exposure among sex workers.

Table 2

Syphilis prevalence among female sex workers recruited by time–location sampling in Mongolia by survey site (n=761)

Syphilis and risk factors

Seventy two per cent (150/209) of syphilis-positive participants were younger than 35 years compared with 64% (351/552) of syphilis negative participants (p=0.009, table 3). Syphilis-positive participants were, in general, of lower educational attainment than their negative counterparts.

Table 3

Selected characteristics of female sex workers recruited by time–location sampling in Mongolia by syphilis status (n=761)

Evaluation of sexual behaviour indicators by syphilis status disclosed few differences. Statistically significant association was observed for client occupation, with 66% of syphilis-positive participants (n=138/209) reporting blue collar working class clients compared with 54% (n=299/552) of syphilis-negative participants (p=0.003). Genital discharge and ulcers were more often reported by syphilis-positive than syphilis-negative FSWs. Genital ulcers were reported by 11% (22/209) of positive participants compared with 6% (31/552) of negative participants (p=0.018). More syphilis-positive (67%, 132/198) than syphilis-negative FSWs (56%, 297/526) were tested for HIV in the past 12 months (p=0.013).

In multivariate analyses (table 4), among participants aged ≤30 years, each 1 year increase in age was associated with 5% (OR=1.05, 95% CI 0.99 to 1.11) increased odds of syphilis infection, while, among participants over 30 years of age, the same odds decreased by 4% (OR=0.96, 95% CI 0.92 to 0.99) with each year. A trend for reduced odds of infection (up to 60% decrease) with higher education level was seen. FSWs from Khuvsgul province had the lowest odds of being syphilis positive compared with FSWs from Ulaanbaatar (weighted OR=0.39, 95% CI 0.19 to 0.81). Having blue collar worker clients was associated with syphilis status in the weighted analysis (weighted OR=1.73, 95% CI 1.09 to 2.76). Occurrence of genital ulcer in the past 12 months increased the odds of testing positive for syphilis more than two times (OR=2.24, 95% CI 1.17 to 4.28). Consistent condom use was associated with decreased odds of infection, but the reduced level did not reach statistical significance.

Table 4

Multivariate logistic regression of relationship between risk factors and syphilis status of female sex workers recruited via time location sampling in Mongolia (n=761)


As with previous rounds of SGSS, this survey found no HIV (0%) but a high prevalence of syphilis (27.5%) among FSWs in selected sites of Mongolia. The finding is reflective of overall low HIV and high rates of STI in Mongolia. The 27.5% prevalence found in this survey is greater than the 18.3% syphilis rate reported in the 2009 SGSS.5 The prevalence exceeded 30% in Ulaanbaatar, Darkhan and Dornod sites, which is higher than the corresponding rates seen in the 2009 SGSS. The rates were comparable with the results of a previous survey, remaining around 20% for Orkhon and Khuvsgul provinces. Although, in 2009, efforts were taken to identify and exclude syphilis-treated cases from those with active syphilis, the new high rate is consistent with the increase in syphilis notifications observed in the country over the past decade4 ,6 and hence, probably, indicates a true increase in the distribution of syphilis among FSWs in Mongolia.

The results are likely to be representative of the subgroup of sex workers that was most accessible to the survey team. The survey relied on outreach workers (former or current sex workers) to correctly identify and enrol FSWs. It is possible, that they only recognised and enrolled FSWs based on their own experience and knowledge of FSWs in the area. Limited resources and time did not allow comprehensive mapping and formative research of all potential venues, forcing the survey team to resort to sampling of venues known to have FSWs. Most participants were recruited at open venues or streets. FSWs recruited by NGOs were mainly from saunas and massage parlours, which were the type of venues hard to access. These participants, however, comprised only 6% of all participants, and a sensitivity analysis excluding them did not produce meaningful changes to the results. Thus, it is likely that the sample might have been skewed towards FSWs who were more visible and relatively easy to access with high syphilis and low HIV prevalence rates.

Absence of HIV among FSWs is in sharp contrast with the national HIV surveillance data, where 11 (39.3%) of 28 female cases registered to date have been identified as sex workers. However, review of epidemiological data of these 11 cases showed that all except one had been engaged in commercial sex activities abroad. Altogether these facts indicate that there are other hidden networks of FSWs with a different risk profile that the survey did not reach.

Limited transmission of HIV among FSWs might be explained by the lack of HIV introduction into the sexual networks of FSWs and minimal overlap with men who have sex with men (MSM), who constitute the majority of HIV cases reported in the country. The parallel survey conducted among 200 MSM recruited using respondent-driven sampling (RDS) found an HIV prevalence of 10.7% (RDS-weighted 7.5%) and syphilis prevalence of 4.1% (weighted 3.4%). Only 1% of MSM reported having sex with FSWs in the past 12 months.16 Injection drug use appears to be limited and no cases of HIV transmission occurring through drug use have been reported.

Despite the above limitations, the findings provide important information on the prevalence of HIV and syphilis among FSWs in Mongolia. Sociodemographic and behavioural characteristics of participants reveal the low socioeconomic status of FSWs, of whom over a third had less than complete secondary education. STI symptoms were common, suggesting a potentially high prevalence of STIs other than syphilis among FSWs. High level of condom use at last sex (over 80%) and reasonable HIV knowledge and testing (nearly 60%) levels were seen. However, consistent condom use and coverage by HIV/STI prevention programmes were about 50%, which together with observed high prevalence of syphilis indicate the potential for HIV transmission.

With the dramatic increase in population mobility and the growing number of migrant workers following recent economic growth fuelled by large-scale mining, transport and infrastructure projects in the country, there is a risk of increased vulnerability to HIV and further escalation of an STI epidemic through interaction among the migrant workers, local communities and sex workers. Renewed HIV/STI prevention and treatment efforts among FSWs that consider legal, policy and changing economic environments are needed.

Key messages

  • Although no case of HIV was found, the prevalence of syphilis was 27.8%, demonstrating a continuing syphilis epidemic among female sex workers (FSWs) in Mongolia.

  • High syphilis and suboptimal consistent condom use and prevention coverage indicate the potential for HIV transmission among FSWs if introduced into their sexual networks.

  • With the increase in migration of mining workers, there is risk of increased vulnerability to HIV and further escalation of an STI epidemic.


Supplementary materials

  • Abstract in Mongolian

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

    Files in this Data Supplement:


  • SM and BA contributed equally to the manuscript.

  • Handling editor Jackie A Cassell

  • Contributors SM, BJ, IA, TS, SB and OA contributed to the conception and design of the study. SM, IA, ZS, JB, contributed to conducting the survey. BJ and BA performed data management. BA carried out statistical analyses and wrote the initial draft. SS, SB and TS provided the main directions for data analysis and report writing. All authors contributed to interpretation of the data and writing of the manuscript.

  • Funding Support for this research was provided by the Global Fund to Fight AIDS, TB and malaria grant MON-H-MOH to Ministry of Health, Mongolia.

  • Competing interests None.

  • Patient consent Obtained.

  • Ethics approval Ethical committee of the Ministry of Health.

  • Provenance and peer review Not commissioned; externally peer reviewed.