Report No: ACS11311 Middle East and North Africa Epidemiology and Economics Intelligence to Inform Policy Decisions on Resource Allocation for HIV/AIDS Programs Integrated Biological and Behavioral Surveillance Survey (IBBSS) Among Female Sex Workers and Long Distance Truck Drivers In Djibouti, 2014 April 30, 2014 GHNDR MIDDLE EAST AND NORTH AFRICA Standard Disclaimer: This volume is a product of the staff of the International Bank for Reconstruction and Development/ The World Bank. The findings, interpretations, and conclusions expressed in this paper do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Copyright Statement: The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The International Bank for Reconstruction and Development/ The World Bank encourages dissemination of its work and will normally grant permission to reproduce portions of the work promptly. For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA, telephone 978-750- 8400, fax 978-750-4470, http://www.copyright.com/. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA, fax 202-522-2422, e-mail pubrights@worldbank.org. The 2014 Integrated Biological and Behavioral Surveillance Survey (IBBSS) was implemented by the Executive Secretariat for Tuberculosis, HIV/AIDS and Malaria, Djibouti. FHI 360 provided technical assistance through the ROADS II project, a U.S. President’s Emergency Plan for AIDS Relief (PEPFAR)- funded project in providing support to the Government of Djibouti’s HIV/AIDS prevention, care and support programs, as well as its HIV surveillance efforts. Funding for the IBBSS was provided by the World Bank and U.S. Agency for International Development (USAID) and PEPFAR. Recommended citation: Government of Djibouti. Integrated Biological and Behavioral Surveillance Survey (IBBSS) among Selected Female Sex Workers and Long-Distance Truck Drivers in Djibouti 2014 CONTENTS LIST OF FIGURES.......................................................................................................................... iii LIST OF TABLES ........................................................................................................................... iv ACRONYMS.................................................................................................................................. v EXECUTIVE SUMMARY ................................................................................................................. 1 ACKNOWLEDGMENTS .................................................................................................................. 6 1 INTRODUCTION/BACKGROUND .............................................................................................. 7 2 METHODOLOGY ................................................................................................................... 11 2.1 Phase I: PSA, Mapping and Estimation of FSWs and LDTDs .............................................. 11 2.1.1 Estimated Number of FSWs by Locations ........................................................... 11 2.1.2 Long-Distance Truck Drivers ............................................................................... 12 2.2 Phase II: Pilot Survey ......................................................................................................... 13 2.3 Phase III: Implementation of Main IBBSS Survey .............................................................. 13 2.3.1 Sample Size ......................................................................................................... 13 2.3.2 Female Sex Workers ........................................................................................... 14 2.3.3 Truck Drivers ....................................................................................................... 14 2.3.2 Sampling Procedures .......................................................................................... 15 2.3.3 Ethical Issues ....................................................................................................... 16 2.3.4 Recruitment and Training of Personnel .............................................................. 16 3 DATA COLLECTION ............................................................................................................... 17 3.1 Data Collection Instruments .............................................................................................. 17 3.2 Data Collection Flow and Interview Procedures ............................................................... 17 3.3 HIV/Syphilis Counseling and Testing ................................................................................. 18 3.3.1 HIV Testing .......................................................................................................... 18 3.3.2 Syphilis Testing ................................................................................................... 19 4 DATA MANAGEMENT AND ANALYSIS ................................................................................... 20 4.1 Field Supervision ............................................................................................................... 20 4.2 Data Cleaning .................................................................................................................... 20 4.3 Data Analysis ..................................................................................................................... 20 5 LIMITATIONS OF THE SURVEY ............................................................................................... 22 6 RESULTS .............................................................................................................................. 23 6.1 Refusal to Take the HIV Test ............................................................................................. 23 6.2 Female Sex Workers .......................................................................................................... 23 6.2.1 Socio-Demographic Characteristics of FSWs ...................................................... 23 6.2.2 Period of Work as an FSW .................................................................................. 24 6.2.3 Mobility among FSWs ......................................................................................... 24 6.2.4 Sex History among FSWs .................................................................................... 25 6.2.5 Transactional Sex among FSWs .......................................................................... 25 6.2.6 Sexual Partners among FSWs ............................................................................. 25 6.2.7 Family Planning ................................................................................................... 25 6.2.8 Condom Use between FSWs and Their Clients................................................... 26 6.2.9 Condom Breaks among FSWs ............................................................................. 27 6.2.10 STI/HIV/AIDS Knowledge, Attitudes and Behavior ............................................. 28 6.2.11 Prevalence of Syphilis among FSWs ................................................................... 31 6.2.12 HIV Prevalence among FSWs .............................................................................. 34 6.2.13 Gender-Based Violence ...................................................................................... 38 i 6.2.14 Mapping of FSWs Hotspots in Relation to Facilities and Resources in the Survey Area ................................................................................................................... 39 6.3 Long-Distance Truck Drivers .............................................................................................. 39 6.3.1 Socio-Demographic Characteristics of LDTDs ..................................................... 39 6.3.2 Period of Work as an LDTD ................................................................................. 41 6.3.3 Mobility among LDTDs........................................................................................ 41 6.3.4 Sex History among LDTDs ................................................................................... 41 6.3.5 Where Do LDTDs Pick Sex Partners? .................................................................. 41 6.3.6 Condom Use among LDTDs ................................................................................ 42 6.3.7 Knowledge of STI Symptoms and Infections among LDTDs ............................... 44 6.3.8 Prevalence of Syphilis among LDTDs .................................................................. 44 6.3.9 Knowledge of HIV/AIDS and Its Prevention Strategies ....................................... 46 6.3.10 HIV Prevalence among LDTDs ............................................................................. 48 7 DISCUSSION AND CONCLUSIONS .......................................................................................... 52 8 RECOMMENDATIONS ........................................................................................................... 57 9 APPENDICES ........................................................................................................................ 59 Appendix 1. Statistical Tests for HIV and Syphilis prevalence by Various Characteristics among FSWs 59 10 REFERENCES AND FURTHER READING ................................................................................... 61 ii LIST OF FIGURES Figure 1. HIV Prevalence among FSWs by Age ..................................................................................... 34 iii LIST OF TABLES Table 1. Number of Hotspots and Estimated FSWs by Area and Time ................................................. 12 Table 2. Number of Trucks at PK12 ....................................................................................................... 13 Table 3. Achieved Sample Size .............................................................................................................. 15 Table 4. Socio-Demographic Characteristics of the FSWs .................................................................... 23 Table 5. Family Planning and Contraceptive Use among FSWs by Socio-Demographic Characteristics....................................................................................................................................... 26 Table 6. Condom Use among FSWs with Occasional and Regular Clients in the Last Sexual Intercourse ............................................................................................................................................ 27 Table 7. Condom Breakage among FSWs ............................................................................................. 27 Table 8. Percentage of FSWs Who Ever Heard of an STI and Who Could Describe Any Symptom in Women and Men .............................................................................................................................. 28 Table 9. Self-Reported Prevalence of STIs and Care/Assistance Sought by FSWs by Socio- Demographic Characteristics ................................................................................................................ 30 Table 10. Prevalence of Syphilis among FSWs ...................................................................................... 31 Table 11. Experience on GBV ................................................................................................................ 39 Table 12. Socio-Demographic Characteristics of Truckers.................................................................... 40 Table 13. Condom Use with Paid and Non-Paid Female Sex Partners among LDTDs .......................... 43 Table 14. Percentage of LDTDs Who Ever Heard of an STI and Could Describe Any Symptom in Men44 Table 15. Prevalence of Syphilis among LDTDs ..................................................................................... 45 Table 16. Percentage of LDTDs Reporting Knowledge of HIV/AIDS...................................................... 47 Table 17. HIV Testing and Awareness of HIV/AIDS and Drugs That Help Treat People Who Have AIDS among Truckers by Socio-Demographic Characteristics .............................................................. 48 Table 18. HIV Prevalence among LDTDs ............................................................................................... 49 iv ACRONYMS AIDS Acquired Immunodeficiency Syndrome ART Antiretroviral Therapy ES Executive Secretariat for AIDS, Tuberculosis and Malaria FGD Focus Group Discussion FP Family Planning FSW Female Sex Worker GBV Gender-Based Violence GIS Geographic Information System GORD Government of the Republic of Djibouti HTC HIV Testing and Counseling HIV Human Immunodeficiency Virus IBBSS Integrated Biological and Behavioral Surveillance Survey IOM International Organization for Migration LDTD Long-Distance Truck Driver M&E Monitoring and Evaluation MARP Most-at-Risk Populations MOH Ministry of Health NGO Nongovernmental Organization ODK Open Data Kit PSA Pre-Survey Assessment RA Research Assistant RH Reproductive Health ROADS Roads to a Healthy Future Project STI Sexually Transmitted Infection SW Sex Worker TB Tuberculosis TLC Time-Location Cluster U.S. United States UNAIDS Joint United Nations Programme on HIV/AIDS USAID U.S. Agency for International Development WHO World Health Organization v EXECUTIVE SUMMARY T his survey, the first of its kind in Djibouti, was based on linked anonymous HIV, syphilis and behavioral surveys to examine the magnitude and risk factors for HIV infection among female sex workers (FSWs) and long-distance truck drivers (LDTDs). The objectives of this survey were: 1. To identify/validate locations where FSWs and LDTDs operate or where they are accessible in Djibouti Ville 2. To assess baseline behaviors among FSWs and LDTDs in Djibouti Ville related to HIV/AIDS, sexually transmitted infections (STIs), reproductive health (RH) and family planning (FP) 3. To measure the prevalence of HIV among FSWs and LDTDs 4. To measure the prevalence of syphilis among FSWs and LDTDs 5. To assess the extent of health service access and needs among FSWs and LDTDs 6. To provide recommendations for evidence-based interventions among survey populations in Djibouti. Methods To select eligible LDTDs for the survey, a two-stage time-location cluster (TLC) sampling approach was adopted. A “take all” approach was adopted for FSW. Socio-demographics and behavioral information was collected via structured individual interviews using Samsung electronic tablets. Blood samples were collected and linked to an individual’s socio-demographics and behavioral data. Biological indicators from both FSWs and LDTDs included testing for recent and untreated syphilis using a non-treponemal test and HIV prevalence testing using a Determine kit. Those who tested HIV sero-positive were confirmed using an Immunocomb kit; those who tested positive for syphilis were confirmed using Treponema Pallidum Hemoagglutination Assay. The protocol, consent forms and draft questionnaires were approved by FHI 360’s Committee for Protection of Human Subjects, with the approval of the Ministry of Health (MOH) of Djibouti. Sex Workers: Key Findings HIV/STI Prevalence Overall, the prevalence of HIV among FSWs was 13%—six times higher than that of the general population of Djibouti, which is estimated at 2.5%. The HIV prevalence was lowest for those aged 18–24 years (5%) and highest among those aged 40 years and over (36%). The syphilis prevalence among FSWs was (5%), five times higher than the national estimated prevalence of 1%. Basic characteristics  A majority (52.5%) of the FSWs operated from service bars, while 22% operated from homes and the remaining 25% operated from the streets or brothels.  A majority (68%) of the FSWs were under the age of 30, and the mean age in this population was 27.1.  The illiteracy levels ranged between 41.7% and 85.2%, depending on the survey site. 1  Almost half (49.9%) of FSWs were unmarried and living alone. Another 29% of the FSWs were either divorced or separated. Sexual and Reproductive Health  More than 60% of FSWs from all the survey sites used contraceptives, and more than 72% of the FSWs were using contraceptives at the time of recruitment. More than 40% of the FSWs had had unplanned pregnancies in their lifetime.  The condom was the most preferred contraceptive method (50%), followed by daily hormonal pills (22%) and injectable hormones (17%). Between 4% and 6% of those who used contraceptives used a combination of at least two contraceptive methods. Sexual Patterns and Behaviors  Of the 363 FSWs recruited, 252 (69%) had had their first sexual encounter when they were aged under 18 years, but 72% had started engaging in commercial sex after attaining the age of 18.  Those who had had an early sex debut were more likely to start engaging in commercial sex at an earlier age than those who had delayed their debut.  About 63% of the FSWs had regular clients, while 82% had occasional clients.  Between 50% and 56% of FSWs (depending on survey site) had regular partners. Condom Use  Between 29% and 40% of all sexual encounters were unprotected when engaging with occasional and regular clients, respectively.  Condom use among FSWs working in brothels was at only 47%, while among those working elsewhere it was 57%.  Only 38% of sex workers (SWs) in brothels used condoms regularly, compared to more than 48% of those working elsewhere.  Incidences of condom breaks within a month before the recruitment drive were at 14%, while 9% reported a condom break during the last sexual intercourse. Knowledge of STIs  At least 164 (45%) of the FSWs reported having heard of STIs in their lifetime. Of these, 19.0% had experienced vaginal discharges, 16.3% had abdominal pain and 10.4% had genital ulcers. Only 18% of brothel-based FSWs reported having heard of STIs. At least 69% of those who operated from public places such as bars and Chicha beits had heard of STIs, followed by those who operate from homes (60%).  At least 50% of those with STI-like symptoms sought assistance from public health clinics/ hospitals, while 23% sought assistance from private pharmacies. Another 19% took home remedies, while 9% consulted private clinics and/or hospitals. At least 40% of those aged over 40 and those aged 30–34 years (44.0%) preferred to buy medicine from private pharmacies. Gender-Based Violence  At least 74 FSWs reported cases of GBV, and 28% of these were of sexual nature.  Close to 60% of FSWs who experienced GBV did not consider themselves at risk of HIV or STIs. 2  HIV prevalence among the 74 women who reported to have experienced GBV was 10%.  About (39%) of perpetrators of GBV were clients and friends were implicated in 11% of cases. Recommendations Based on these findings, we recommend the followings:  Demand for STI diagnosis and treatment among the SW population in the survey area should be increased by improving their knowledge on sexual health, the major symptoms of STIs, and the link between untreated STIs and HIV.  While consistent condom use (i.e., every time) should constitute the major intervention strategy with both paying clients and non-paying partners, prevention programs should place more emphasis on promoting consistent condom use with both categories.  FSW-friendly, outreach and mobile STI, condom promotion and voluntary counseling and testing services should constitute the priority intervention.  Further research to understand the spread of STIs and risk factors for STIs among SWs is warranted. In this survey, blood samples were tested only for active syphilis, although this was supplemented by self-reported STIs.  This survey provides baseline data that have not been available on various issues related to HIV and syphilis infections and prevalence in Djibouti. This report provides critical data that will be useful in formulation of other related studies in future. There is need for further research on areas and sub-populations not covered by this survey such as other clients of sex workers (dock workers, military etc) and other vulnerable populations such as youth. We anticipate that a follow-on survey to evaluate interventions that will take between 2 and 5 years to implement will be designed based on findings of this survey. Qualitative research will also provide an in-depth understanding of emerging issues.  It is important to note that both groups largely comprised of young participants (less than 25 years old). Considering that this survey only recruited adults (above the age of 18 years), there are possibilities that children (under the age of 18 years) may also be working as sex workers and truckers. Future studies and programs should also focus on these groups and find strategies to reach to younger ages through media including social media Long-Distance Truck Drivers: Key Findings HIV/STI Prevalence  Among the 526 truckers tested, 1% tested HIV positive and 1.5% were ere positive for syphilis.  Interestingly, 4 of the 5 HIV-positive cases reported among LDTDs were found among the “literate.”  Similar to the findings on FSWs, there was a correlation between anal sex and acquisition of HIV. Basic characteristics  More than 50% of the truckers were younger than 30 years, with a mean age of 31 years, and only 16% were aged 40 years or over. A majority of truckers (94%) slept in their trucks during the road trips, while only 5% and 2% slept either in a home or in a hotel, respectively. 3  Literacy levels were relatively high among the truckers (64%) being able to read and write.  Close to 45% of all LDTDs were currently married.  About 63% of the truckers had worked for less than 3 years, indicating that the majority were still new in their jobs. Sexual Patterns and Behavior  More than 77% of the LDTDs had their first sexual encounter when they were between the age of 16 and 20 years.  Those who had had a sex debut at an age under 15 years were more likely to seek a paid sexual encounter than those who had had a delayed debut.  More than 77% first sexual encounters among the LDTDs were with a non-paid female.  More than 10% of the LDTDs had paid for sex in the last one month before the survey, and another 25% of having had sex with a non-paid partner. Condom Use  Only 3.3% of the 526 LDTDs reported experiencing a condom break.  The rate of consistent condom use with occasional and regular partners was only 72%. The rate of those who reported having had sex with paid female sex partners was 20%. More than 90% indicated that they had used a condom the last time they had sex with paid partner, and 76% had used condoms with non-paid female sex partners. Knowledge of STIs and STI Symptoms and HIV and Self-Perception of Risk to HIV/STIs  Knowledge of symptoms of STIs was relatively low, with only 6.5% reporting to have heard of STIs.  The prevention methods for HIV/STIs that were correctly identified by most LDTDs unassisted included safe blood transfusion (92%), always using a condom while engaging in sexual intercourse (89%), having sex with only one uninfected female partner (88%), avoiding sharing of injection equipment (86%) and abstaining from sexual intercourse (85%).  Some misconceptions of HIV/STIs and transmissions were noted. About 31% of truckers who were aware of HIV/AIDS felt that one can prevent HIV infection by cleaning genitals after sex, by avoiding being bitten by mosquitoes or other insects (38%) and by not sharing clothes or utensils (17%). Almost 72% of the LDTDs believed that it is possible to guess if someone is HIV positive through visual assessment.  A high percentage (75%) reported having been previously tested for HIV. A majority (93%) of those tested indicated that the last test was voluntary, and 75% were counseled on HIV/AIDS and its prevention methods.  Slightly more than 30% of the LDTDs were aware of drugs that can be used for the management of HIV/AIDS, and 16% had specific knowledge of antiretroviral therapy (ART).More than 50% of the truckers knew of someone with AIDS or someone who died from AIDS. 4 Possible Risk Factors for Syphilis and HIV  About 18% of the LDTDs felt that they were at risk of contracting HIV, but 57% did not consider themselves to be at risk.  A small percentage (less than 10%) considered themselves at risk of HIV by engaging in anal sex, having more than five sex partners, and having had a condom break during sex. Recommendations Based on these findings, we recommend the following.  There is a need to increase the demand for testing of HIV and syphilis status among LDTDs.  Future studies should investigate whether LDTDs, the majority of whom are foreigners, face language and cultural barriers that hinder their access to critical information on HIV and syphilis prevention.  There is a need to design strategies to improve the knowledge of LDTDs on STI symptoms and self-assessment of risk of HIV and syphilis infections.  A significant proportion of LDTDs indicated that they constantly engage in sex with unpaid sex partners, a population that was not included in the present survey. Future studies should investigate the role of such sexual associations in HIV/STI epidemics.  LDTDs have traditionally been labeled as a special vulnerable group and a transmitter group for HIV and STIs. However, this survey found that the prevalence of HIV and syphilis among this group were below the national averages of Djibouti and Ethiopia, the native country of most LDTDs. Future studies should investigate the reasons behind these apparent discrepancies.  Majority of LDTDs who participated in this survey were from Ethiopia. There is therefore a need to launch cross-border HIV/STI prevention and sensitization programs targeting such highly mobile and vulnerable groups.  5 ACKNOWLEDGMENTS The Ministry of Health (MOH) acknowledges the efforts of all that contributed to the successful conduct of the 2014 Integrated Bio-Behavioral Surveillance Survey (IBBSS) among female sex workers (FSW) and long-distance truck drivers (LDTD) in Djibouti. The contribution of the Executive Secretariat (ES) for HIV, TB and Malaria was instrumental and highly appreciated in the successful completion of the survey. Also worthy of commendation are the contributions of the MOH/ staff, Executive Secretary (Executive Secretariat (ES) for HIV, TB and Malaria), Technical Director (Executive Secretariat (ES) for HIV, TB and Malaria), PLSS Coordinator, Director Laboratory Services and the efforts of the survey team members including the coordinator, supervisors, data manager, interviewers, counselors and laboratory technicians in the generation of high quality data are highly appreciated. We also appreciate the survey participants for agreeing to be part of the survey and responding to the behavioral questions as well as giving blood for HIV and syphilis testing. Our special appreciation goes to members of the survey technical committee, whose technical oversight functions guided the survey. The cooperation and support of health care providers and in-charges of the health facilities involved in the survey are also highly appreciated. We acknowledge FHI 360 for the technical leadership on this survey working closely with different GORD technical staff including MOH, ES, PLSS, Laboratory services among others. The Government of Djibouti acknowledges the financial and technical support of the World Bank and the U.S. Agency for International Development (USAID) through FHI 360/Roads to a Healthy Future (ROADS II) Project for this survey. Findings presented in this report will assist in advocacy and program design and planning towards appropriate and improved STI HIV/AIDS and STI interventions among key populations in Djibouti. Secretary General; Ministry of Health Djibouti 6 1 INTRODUCTION/BACKGROUND 1.1 HIV/AIDS in Djibouti Djibouti lies in northeast Africa on the Gulf of Aden, at the southern entrance to the Red Sea. It borders Ethiopia, Eritrea and Somalia. The country is mainly a desert with scattered plateaus and highlands. The country has a population of approximately 850,000 people and a 2011 estimated per capita gross national income of US$1,513 (UN Data: World Statistics Pocketbook 2013). The 2013 CIA World Fact Book estimates that 42% of the population lives below the poverty line. The estimated adult HIV prevalence of 2.5%, the highest among Arab League countries. HIV has reached epidemic levels in Djibouti: An estimated 16,000 people currently live with the virus. More than 50% (about 9,000) of those who are HIV positive are women, and more than 1,000 are children. However, recent reports1 show that HIV prevalence in Djibouti has been relatively stable in the last 10 years (UNAIDS 2012). Djibouti’s current Strategic Plan for the Fight against HIV 2012–2016 has identified SWs and their clients, especially the LDTDs, dockworkers and other migrant populations, as highly vulnerable groups. SWs and their clients account for 21% of new HIV infections (UNGASS 2012). This is partially because there is significant stigma and discrimination associated with HIV in Djibouti, which inhibits testing and disclosure of sero-positive status, which in turn jeopardizes access to prevention, care and treatment services. Due to the existing knowledge gaps on the dynamics of HIV epidemics in Djibouti, there is a paucity of reliable data and information that can be used to formulate policies and prevention strategies to halt the spread of HIV and other STIs among vulnerable groups. There is therefore a need to generate reliable and up-to-date data that the Government of the Republic of Djibouti (GORD) and its development partners can use to make accurate and informed decisions on HIV/AIDS/STI prevention and management strategies, especially among vulnerable groups. Lack of accurate and up-to-date data also undermines the efforts of GORD to mobilize resources for appropriate responses to the ever-changing HIV transmission and management dynamics. The transport corridors are important avenues for HIV/STI transmission in Djibouti. Various studies have shown a relationship between HIV and mobility. Although mobility and migration are not risk factors for HIV per se, the harsh conditions of travel and long periods of isolation (from family, peers and friends) experienced by LDTDs can force them to engage in behaviors that are strongly associated with increased vulnerability to HIV. These conditions may also create barriers to accessing HIV prevention, treatment and care support. In most cases, the LDTDs seek sexual services from FSWs who patronize the areas where trucks stop, especially those located near major cities. This is particularly worrying because HIV prevalence has been reported to be as high as 20% among FSWs in Djibouti’s capital city (Marcelin, Anne‐ Geneviève, et al. 2002) 1 Key reference documents include Djibouti’s Strategic Plan in the Fight against HIV 2012–2016; Report on Development of a Strategy for Migrants and Interacting Populations in Djibouti, UNAIDS 2012; Vulnerability to HIV in the Context of Cross-Border Migration and Mobility in the Red Sea and Gulf of Aden, UNAIDS 2012; Qualitative Study on the Risk of Adolescent Girls and Young Vulnerable Women to HIV/AIDS in Djibouti, UNICEF 2011; Situation Analysis of Priority Groups for Combination Prevention of HIV/AIDS in Djibouti, UNAIDS 2011; Report on the Study of Modes of Transmission of HIV in Djibouti, UNAIDS 2010; and Study on Knowledge, Attitudes and Practices of Youth in Djibouti, UNICEF 2010. 7 It has been shown that LDTDs, military personnel, sailors and migrant laborers form the bridging populations in HIV transmission, and this group of vulnerable people often record high prevalence of HIV due to their interactions with FSWs. They are among the main clients of FSWs, whose HIV prevalence is known to be much higher than the national average. The LDTDs may therefore serve as a bridging population that could in turn serve as conduits of infection from the key populations (such as SWs) to their regular partners, including spouses and wives. 1.2 Anatomy of the Addis Ababa-Djibouti Transport Corridor The Addis Ababa-Djibouti Transport Corridor is an 850-km two-way stretch that has more than 27 towns/stops. Approximately 4,000 truck drivers and their assistants traverse this corridor every day. A majority of LDTDs stop at PK12, which is the main stop along the corridor in Djibouti. The stop, whose name is derived from the distance to Djibouti Ville, the capital center of Djibouti, is home to some 25,000 people. At least 700–800 trucks that travel the corridor between Djibouti and Addis Ababa park at this point every day; therefore, PK12 constitutes an important focal point for possible interaction between LDTDs, SWs, and the local community (http://www.unicef.org/infobycountry/ djibouti_54193.html). Such concentrated interaction has the potential to exacerbate the HIV epidemic in the region. LDTDs are considered to be at higher risk of contracting and transmitting HIV because of their job-related mobility, which in many countries has shown to lead to an increased patronage of commercial and casual sex partners (Lacerda et al. 1997). Existing data show that a majority of truck drivers along this corridor are Ethiopians. Most truckers stop for meals, to sleep overnight and to meet with FSWs in hotels and bars, and it is not unusual for the truckers to engage in unprotected sex with these FSWs. FSWs may also prefer LDTDs over local residents because the LDTDs may have more money. It has been reported in several studies that FSWs may opt to have sex without condoms for higher pay. This exposes them and their clients to higher chances of infection (Ramjee et al. 2002). These vulnerabilities have been classified in three categories: individual, social/economic and programmatic. Individual vulnerability could arise among LDTDs as a result of the loneliness and isolation that the LDTDs experience during long hours of driving or as a result of separation from their home, families and peers. LDTDs may also experience a disconnection from their normal social behavior and traditions, and differences in culture along the transport corridor and the stopover points. Social and economic vulnerability may be more common among SWs, perhaps as a result of extreme poverty that pushes them to sex work and a lack of legal status and rights, especially if they are foreigners, and as a result of stigma and discrimination associated with transactional sex. In addition, gender inequality makes FSWs vulnerable to violence and abuse, including non-consensual sex and more-risky anal sex. Programmatic vulnerability may arise due to the limited access to health, information and other support services that mobile people like LDTDs and SWs experience. A combination of these vulnerabilities can predispose individuals to engage in risky behaviors associated with high chances of acquisition and transmission of HIV/STIs. Previous studies have shown that there are insufficient health service points and that poor attention is paid to the specific needs of truckers and SWs, especially along the Addis Ababa-Djibouti Transport Corridor. This is particularly worrying because it has been shown that both FSWs and dock workers/truckers engage in a significant amount of risky behavior and that condom use is low or inconsistent. A previous survey showed that sexual health seeking behavior was poor among SWs 8 and LDTDs; nearly 40% of LDTDs and dockworkers and 50% of SWs continue to have sex while experiencing STI-like symptoms. The survey also showed that the current health facilities along the corridor are insufficient and that there are no clear plans to increase their numbers in the future to cope with the huge rise in traffic along this corridor (UNDP (2011). There is also a need to develop strategies that can estimate the number of SWs and LDTDs at different contact points (“hotspots”), estimate the volumes of transactional sex taking place on the major transport corridors and determine if the majority of such sexual encounters are conducted safely by using protective aids such as condoms. There is also a need to determine the availability of health care facilities that provide testing and management support for HIV/AIDS/STIs along the corridors. It is also important to identify the spatial distribution of “vulnerable places.” Places of vulnerability, as defined by the International Organization for Migration (IOM), are those areas where migrants and mobile populations live, work, or pass through, or from which they originate. They may include border posts, ports, truck stops or hotspots along the transport corridors, construction sites, commercial farms, fishing communities, mines, migrant communities and urban informal settlements, migrant sending sites, detention centers and emergency settlements. Within each of these places, there exist micro-hotspots that include bars, hotels, popular eating and drinking spots and discotheques. Transactional sex is openly or covertly available in such spaces and GBV and sexual abuse are not uncommon. 1.3 Rationale There are considerable gaps in the quantity and quality of information available in Djibouti on HIV/AIDS to help understand the underlying dynamics of the HIV epidemic and its likely future course. In this context, in collaboration with the World Bank and U.S. Agency for International Development (USAID)/Djibouti, FHI 360 conducted this IBBSS to generate relevant data to inform HIV-related policy discussions, strategic decision making, mobilization and allocation of resources and measurement of HIV and AIDS program results in the country, with a focus on FSWs and LDTDs. 1.4 Overall Objective The main objective of the current survey was to generate baseline data to inform HIV-related policy discussions, to help in strategic decision making and to support resource mobilization for HIV interventions. The information generated is meant to serve as a baseline for subsequent evaluations of the effectiveness of HIV interventions among FSWs and LDTDs in Djibouti. 1.5 Specific Objectives Below are the specific objectives of this survey that were achieved in three phases (which are discussed later): 1. To identify/validate locations where FSWs and LDTDs operate or where they are accessible in Djibouti Ville 2. To assess baseline behaviors among FSWs and LDTDs in Djibouti Ville related to HIV/AIDS, STIs, RH and FP 3. To measure the prevalence of HIV among FSWs and LDTDs 4. To measure the prevalence of syphilis among FSWs and LDTDs 5. To assess the extent of health service access and needs among FSWs and LDTDs 9 6. To provide recommendations for evidence-based interventions among survey populations in Djibouti 1.6 Survey Area/Sites This survey covered four areas/quartiers of Djibouti Ville: Quartier 2, Quartier 4, Arhiba and PK12. These sites were identified by the GORD for HIV intervention through the Roads to a Healthy Future Project (ROADS). PK12 is the only truck stop within Djibouti Ville and was therefore naturally selected as the site for sampling truckers. Arhiba, Quartier 2 and Quartier 4 were selected because a large number of FSWs in Djibouti Ville operate there. 1.7 Survey Population The survey population included FSWs and LDTDs. For purpose of this IBBSS, an FSW was defined as any female 18 years or over who had received money or other valuable gifts/incentives in exchange for sex in such areas as brothels, bars, restaurants, night clubs, or hotels, or on the street, within the month prior to survey recruitment. An LDTD was operationally defined as a long-haul driver or assistant/loader, aged 18 years or over, plying the Djibouti-Addis Ababa transport or similar route, who stays away from the home area for durations of at least 1 day. 10 2 METHODOLOGY The second generation surveillance approach was used for this survey. This approach was developed by the World Health Organization (WHO) for regular, systematic collection, analysis and interpretation of information for use in tracking and describing changes in the HIV/AIDS and STI epidemic over time. There were three main phases to the research: 1) pre-survey assessment (PSA), mapping and estimation of FSWs and LDTDs; 2) a pilot survey; and 3) the main IBBSS. 2.1 Phase I: PSA, Mapping and Estimation of FSWs and LDTDs The PSA was undertaken to gain a better understanding of the survey groups and relevant contextual factors, to examine how the survey groups can be reached, to build rapport with the target population, to sensitize the population on the survey, to identify potential challenges or safety concerns for either the survey team or survey population, and to assess the population’s receptivity to the survey. This activity was also undertaken to get a rough estimate of the number of FSWs and LDTDs operating from the survey areas. We anticipate that more data will be collected in the future using other methods of size estimation to triangulate the PSA results and validate the estimated number of FSWs and LDTDs in PK12, Quartier 2, Quartier 4 and Arhiba. The PSA was conducted through focus group discussions (FGDs) with FSWs and LDTDs to get a better understanding of their operations—where they can be found and the time of the day when the most or least number of FSWs are available at the hotspots—and in-depth interviews with key opinion leaders, including local administrators, FSW welfare association leaders and government officials, to get additional information and corroborate information from the FSWs and LDTDs. The data collectors also observed and counted the number of FSWs at specific times of the day. The main topics selected were the nature and characteristics of the FSWs and LDTDs, their work, their estimated numbers at the hotspots, the types of clients for the FSWs, the mobility pattern of the target groups and their health care-seeking behavior. This phase targeted 8–12 focus groups of SWs (2–3 in each of the four quartiers of Djibouti Ville), with about 10 participants each, and 2–3 focus groups of LDTDs, with 10 participants each. At least 12 in-depth interviews were conducted with opinion leaders, 3 in each quartier. (A separate, detailed pre-survey assessment report is available). Mapping of FSWs and LDTDs was conducted to build comprehensive sampling frames from which the representative samples were to be drawn. The exercise targeted all possible locations in Djibouti Ville (PK12, Quartier 2, Quartier 4 and Arhiba). The locations were identified through community leaders, SWs, peer educators and other individuals knowledgeable of the areas. As most transactional sex in these areas occurs in bars and lodges, these were also mapped and covered in the survey. For each mapped bar or lodge, some key information, such as operational days, operational time, and estimated size, was collected through interviews with staff members. 2.1.1 Estimated Number of FSWs by Locations The numbers presented in Table 1 are from the mapping exercise conducted by trained research assistants (RAs), who were paired with one FSW peer educator in each site to assist in the identification of FSW hotspots and in the gathering of information, such as site characteristics, estimated number of FSWs and working hours. After a peer educator showed the RA a hotspot, the RA approached all females present and directly asked if they were engaged in sex work. Additional 11 data were collected through FGDs with selected FSWs to corroborate the number established from the direct count. The FGDs revealed that there is a significant number of young “hidden” FSWs who are in contact with their clients through telephone and/or social media; this hidden population was therefore not found at the hotspots. Table 1. Number of Hotspots and Estimated FSWs by Area and Time Number of FSWs at hotspots by time Estimated population Number of Site of quartier hotspots Maximum Medium Minimum Quartier 2 6,426 11 230 134 72 Quartier 4 12,403 8 101 60 40 Arhiba 12,000 9 105 68 35 PK12 19,117 18 125 65 31 Total 49,946 46 561 327 178 In the four survey sites, 46 hotspots were identified, including Chicha beits, informal “alcohol selling houses,” bars/hotels/restaurants, brothels, lodges and street corners. An RA accompanied by a peer educator visited each hotspot three times a day, every day of the week. The highest number and the lowest number of FSWs during all these visits were defined as the maximum and minimum counts, respectively. The average was considered the medium count. It was estimated that a maximum of 561 FSWs operated from these hotspots during peak hours (time that the highest number of FSWs can be found at the hotspot, when they are likely to find clients), 327 at medium hours (in-between time: no highest or lowest number of FSWs can be found at the hotspot, when not so many clients frequent the hotspot) and 178 at lean hours (time during which the least number of FSWs can be found at the hotspot, when there are few clients, in most cases coinciding with regular working hours). Quartier 2 had the highest number of FSWs, followed by PK12, Arhiba and Quartier 4. Quartier 2. Hotspots in Quartier 2 consisted of mainly Chicha beits, alcohol dens, khat dens and/or brothels. Quartier 4. The eight hotspots in Quartier 4 included Chicha beits, lodges, brothels and street corners. Arhiba. In Arhiba, hotspots included brothel-like houses, street corners and Chicha beits. PK12. Most of the hotspots in PK12 were Chicha beits (14) and informal “alcohol selling houses” (3). There was one bar/hotel. It was reported that there had been several police operations targeting FSWs in the area, leading to closure of numerous venues where sex work was occurring. 2.1.2 Long-Distance Truck Drivers The number of LDTDs found daily at PK12 was estimated through a review of Customs records and corroborated with truck drivers’ association records. Table 2 shows the estimated number of trucks at PK12 per day based on truck drivers’ association records. The daily average number of trucks indicated by customs in an in-depth interview was 500, which is close to the number obtained from the association’s records. These records indicated that 442 were recorded on Wednesday and 568 were recorded on Thursday. From key informant interviews, it was apparent that each truck has two 12 operators: a driver and an assistant. Therefore, on any given day, there was an estimated 884–1,136 LDTDs (drivers and their assistants) at PK12. The number may be higher, given that some trucks that ply the route do not necessarily stop overnight at PK12. Table 2. Number of Trucks at PK12 Day Date 7 pm Monday Nov 18, 2013 469 Tuesday Nov 19, 2013 440 Wednesday Nov 20, 2013 442 Thursday Nov 21, 2013 568 Friday Nov 22, 2013 525 Saturday Nov 23, 2013 565 Sunday Nov 24, 2013 498 2.2 Phase II: Pilot Survey A 1-week pilot field trial was conducted in two of the survey areas (PK12 and Quartier 2) involving testing various aspects of the survey process, including the survey protocol, methodology, training and data collection. A total of 20 LDTDs were identified and interviewed in PK12, and 26 FSWs were interviewed in Quartier 2. A team of 2 RAs (data collectors) led by a co-investigator conducted the pilot survey to assess the feasibility of the main IBBSS. The research team gathered information to inform improved quality and efficiency of conducting the main IBBSS. The pilot survey was designed to provide vital information on the feasibility of the proposed procedures and, therefore, was a good foundation for collection of high-quality data. Data were collected on HIV knowledge, attitudes and risk behaviors, as well as HIV status. 2.3 Phase III: Implementation of Main IBBSS Survey The final phase of the survey involved conducting a cross-sectional IBBSS among the survey populations (FSWs and LDTDs). 2.3.1 Sample Size Sample sizes for each population (sub-group) included in the IBBSS were calculated on the basis of the following factors typically used in surveys with probability samples (see Section 2.3.2 on sampling):  Expected baseline value of key behavioral indicators (e.g., consistent condom use with various partner types)  Magnitude of change it is desired to be able to detect  Confidence level  Statistical power  Design effect The following formula was used to determine the sample size for target groups for the IBBSS: 13 nD  2 P (1  P ) Z1  P1 (1  P1)  P2 (1  P2 ) Z1   2 2 where: D = design effect P1 = estimated proportion at the time of the first survey P2 = proportion at some future date, such that the quantity (P2 − P1) is the size of the magnitude of change it is desired to be able to detect P = (P1 + P2) / 2 2 2 ∆ = (P2 – P1) Z1- = the z-score corresponding to the probability with which it is desired to be able to conclude that an observed change of size (P2 − P1) would not have occurred by chance Z1- = the z-score corresponding to the degree of confidence with which it is desired to be certain of detecting a change of size (P2 − P1) if one actually occurred For the Djibouti IBBSS, the following assumptions have been made regarding these parameters: 1. Expected baseline value of 50%. Measurements require the highest sample size to detect change when the baseline is 50%, hence this figure was used. If it can safely be assumed that baseline values of all indicators are significantly lower or higher, then sample sizes could be lowered. 2. Desired change to detect: 10%–15%. This refers to the amount of change that can be detected between two survey rounds. For example, if condom use changed by an absolute 10%–15%, this would be detected as a statistically significant change. A lower absolute change would not be detected as statistically significant. Smaller differences require larger sample sizes. 3. The alpha (α) level has been set at 0.05, corresponding to 95% confidence in the observed estimates. 4. The beta (β) level has been set at 0.10, corresponding to 90% power. 5. Design effect of 1.7. This adjusts for the use of sampling designs that are not simple random methods, e.g., cluster sampling. 2.3.2 Female Sex Workers A sample size of 400 FSWs was calculated for this survey. However, based on the pre-survey mapping exercise, the number of FSWs at the four sites ranged from a low of 178 to a high of 561. Therefore, a “take all” approach was adopted, since the numbers involved may not have allowed for any other sampling design (see Section 2.3.2). Of the expected 178–561 FSWs, the actual number recruited for this survey was 363 (see Table 3). 2.3.3 Truck Drivers The above formula derived a sample size of 600 LDTDs at PK12 truck stop. The actual achieved sample size of 526 was finally recruited in this survey. 14 Table 3. Achieved Sample Size PK12 Quartier 2 Quartier 4 Arhiba Total Total FSWs 12 176 54 121 363 Brothel-based 0 10 0 24 34 Service bar/lodge 7 119 30 34 190 Chicha beits and others 1 2 0 30 33 Home-based 4 30 17 31 82 Street-based 0 15 7 2 24 Total LDTDs 526 – – – 526 2.3.2 Sampling Procedures FSWs. The PSA estimated the total number of FSWs in the survey areas (four quartiers) to be between 178 and 561. Therefore, based on this key finding, we decided to adopt the “take all” sampling approach to cover FSWs for the IBBSS. The “take all” sampling attempts to include every eligible and accessible participant within the defined sampling domain using multiple approaches, including field visits, and outreach using peer key informants and community leaders, etc. In this survey, our field teams attempted to access all the eligible FSWs, through community volunteers and liaisons who were working very closely with FSWs on daily basis. Duration of the fieldwork was extended for a longer period to ensure that as many FSWs as possible were recruited or reached in the survey. The survey team managed to recruit a majority of all eligible FSWs in Quartier 2 (see Table 3) and the least number in PK12. PK12 had the lowest number of recruited FSWs (12) out of an expected highest expected (125), followed by Quartier 4 (54 out of an expected total of 101). It should be noted that the number of SWs varied depending on the time of the day, day of the week and season (the number of clients available varied at those times of day and year). The low turnout in PK12 was probably attributed to the nature of the SWs in the area: A majority of them were either young or foreigners who got their clients through phone and/or social media and were therefore not easily available for face-to-face interviews. LDTDs. To select eligible LDTDs for the survey, a two-stage TLC sampling approach was adopted. Based on detailed mapping information on all possible access points for LDTDs within PK12, operational time of the site, average duration of stay at the site, average duration for a return visit at the site, etc., a list of the relevant TLCs for PK12 was developed. For the first stage, a total of 26 field day TLCs, each TLC being of 1-day duration, was covered for the survey. Based on the findings of the PSA, a truck driver was likely to come to PK12 at least once within a period of 26 days. It was calculated that going beyond the 26 days would likely get the same individuals already interviewed and the recall period for participation in the interview was within the month. A systematic random sampling approach covering the selected TLCs was used for selection of the required number of eligible respondents during the second stage. On average, the field team, consisting of six interviewers, recruited about 20 eligible respondents (LDTDs) per day. Respondents were selected systematically from different corners of the TLC. An RA visited the TLC and approached every male present and confirmed that he was a truck driver before requesting an interview. The process was repeated until all eligible participants of the day were covered. The team kept track of total estimated number of LDTDs at the TLC, the response rates and the reasons for refusals. 15 2.3.3 Ethical Issues Participation of all respondents in the survey was strictly voluntary. Measures were taken to ensure that the respect, dignity and freedom of each individual participating in the survey was safeguarded. To guarantee the anonymity of each participant, the names of respondents, their addresses or other identifying information were not included in the questionnaires or on any biomarker tracking forms. The questionnaires were identified with a code. Stickers with numbered codes were used on the questionnaire, on the blood sample vials, on the laboratory reports and on the HIV testing and counseling (HTC) referral forms. Informed consent was obtained from each participant by reading a short paragraph that summarized the survey and the guarantee of confidentiality. After witnessing informed consent, the form was signed by the respondent and the RA. In cases where the respondent did not wish or was not able to provide a signature, the form was signed by the person requesting the consent and by a witness, a person identified by the respondent (usually a fellow truck driver or friend). The witness was deemed valid only if he was a witness after verbal informed consent of the participant had been obtained. During the interview, basic information was provided regarding HIV/AIDS, diagnostic testing, condom use and specific risks for each target population. FHI 360’s Committee on Protection of Human Subjects and Djibouti’s MOH concurrently approved the protocol, consent forms and draft questionnaires of the survey. (Note that the Djibouti National Institutional Review Board was not operation at the time the survey was designed.) 2.3.4 Recruitment and Training of Personnel A field team comprising 1 coordinator, 3 supervisors, 1 data manager and 14 RAs, along with 18 HTC counselors and 18 laboratory technicians provided by the government for the IBBSS process was put in place before the implementation of the survey. The RAs were selected carefully; they had to have a minimum of a secondary school education (10 years of schooling) and had to be experienced in data collection and/or HIV programming. For the testing exercise, the government provided trained and accredited counselors and laboratory technicians. Most importantly, the interviewers and counselors were individuals who were able to read and speak the local language of the interviewee. Two levels of training were implemented: a 2-day training of coordinators, supervisors and data manager, and a 6-day training for RAs, followed by training of HTC counselors and laboratory technicians. At this training, the field team was guided through in-depth sessions on the survey objectives, the methodology and the questionnaires. Their understanding of their roles in the survey, the survey process, ethical considerations and the need for good quality data were also enhanced. Data were collected using electronic devices (tablets). The field teams got a 1-day training on the use of tablets and a 1-day practical session. During the training, each question in the questionnaire was reviewed and role-played and possible challenges identified and addressed. In addition, a 1-day field trial was conducted for the interviewers in two selected sites prior to commencement of data collection. HTC counselors and laboratory technicians were given refresher courses on the processes of counseling, testing of HIV/syphilis among most-at-risk populations (MARPs) and ethics consideration with a special focus on confidentiality. To ensure sustainability and to build the capacity of Djibouti’s institutions, the person responsible for monitoring and evaluation (M&E) in the Executive Secretariat for HIV/AIDS, TB and Malaria (ES) was trained and involved in supervision of the field team. 16 3 DATA COLLECTION 3.1 Data Collection Instruments A standard IBBSS questionnaire developed by WHO, the U.S. Centers for Disease Control and Prevention, FHI 360 and other organizations was modified to suit the Djibouti context and used for data collection. A paper-based questionnaire was used during the pilot survey, but for the main IBBSS survey, the questionnaire for each of the sub-groups was uploaded into Samsung Galaxy Tab 3 tablets for electronic data collection. The questionnaire was translated into French, and the RAs (interviewers) were trained to ask questions without changing and/or distorting the intended meaning and content of the question. Information was obtained on selected behavioral indicators, including sexual history and practices, sexual risk behavior, condom use with different types of sexual partners, knowledge of STIs, STI care-seeking behaviors and FP knowledge and practice, among other issues. Biological indicators from both FSWs and LDTDs included testing for recent and untreated syphilis using a non-treponemal test and HIV prevalence testing using a Determine kit. The blood samples of those who tested HIV sero-positive were confirmed using the Immunocomb kit, while syphilis-positive samples were confirmed using the Treponema Pallidum Hemoagglutination Assay. 3.2 Data Collection Flow and Interview Procedures Data were collected daily using hand-held electronic data collection devices (Galaxy Tab 3 tablets) between March 8, 2014 and April 10, 2014, from FSWs identified at hotspots in the four quartiers, while the LDTDs were identified at the truck park in PK12. Trained RAs conducted face-to-face interviews with the respondents after the HIV testing. HIV testing was conducted by public health care providers (a laboratory technician and a counselor). The following steps were followed during the survey data collection process:  The counselor introduced and explained the survey to the participant.  The counselor obtained informed consent for the biological component from the participant and assigned a unique code (ID) to the client to be used at the laboratory and by the interviewer during the behavioral interview.  The counselor referred and accompanied the respondent to the interviewer.  The interviewer obtained consent from the participant for the behavioral component.  The interviewer conducted the behavioral interview using the participant ID and accompanied the participant back to the counselor.  The counselor conducted pretest counseling and accompanied the participant to laboratory technician.  A laboratory technician drew a blood sample from the participant and labeled the sample with the participant’s survey ID.  A laboratory technician conducted a rapid test and sent results to the counselor.  Blood samples with corresponding code numbers were appropriately stored and those that tested positive for HIV (and/or syphilis for the main survey) were sent to the referral hospital. 17  The counselor provided HIV test results to the participant (if desired) and post-test counseling. Behavioral interviews were conducted in private settings (health facility or community centers) that guaranteed the confidentiality of information provided by the respondent. The survey team in each facility consisted of interviewers, a supervisor, counselors and laboratory technicians. Consent was obtained by a counselor or supervisor who acknowledged that all necessary information was provided and understood. Interviewers ensured that respondents were not interviewed twice by asking respondents if they had been interviewed in the last 30 days and by using a single team for screening of participants. This was done because the target groups were very mobile and the probability of double recruitment at two different survey locations was high. Adequate preparations were made to ensure that the interviews were conducted at a time determined during the pre- surveillance mapping exercise as the most suitable to conduct the interviews and testing. In particular, the exercise was synchronized with the working hours of the FSWs, who normally start working at 4pm. The peak period (when the majority of them were in the hotspots) was determined to be between 4pm and midnight. For security reasons, the field team visited and stayed at the hotspots between 3pm and 9pm as they recruited participants. 3.3 HIV/Syphilis Counseling and Testing Before the behavioral interview, a counselor explained the HIV testing and escorted the participant to the laboratory technician, who drew blood and conducted a rapid test for HIV and syphilis. The counselor escorted a participant for the test only after he/she had voluntarily consented to take the test. Following discussions with the MOH and an analysis of the PSA data, it was agreed that testing would be conducted at the health facility, the ideal place for the required privacy. In PK12, testing was conducted at the ROADS II-supported resource center, after getting approval from the MOH for sufficient privacy and confidentiality. At the other three survey sites, interviews and testing was conducted at the nearest public health facility (the Arhiba and Khor Bourhan health centers). To ensure that there was no interruption of routine service delivery for other members of the public, the field coordinator, supervisors and interviewers made prior visits to officers in charge of the target venues. Most of the participants tested were willing to participate and received their test results immediately after appropriate post-test counseling. Those who did not immediately collect their test results were encouraged to collect them at their convenience. Each participant was given a card with a reference number to enable him/her to collect his/her results whenever he/she wished. Apart from the reference number, no personal identification information was included on the questionnaire or card. The laboratory technician completed the MOH HTC forms/register from which the supervisor extracted the results at the end of each day, linking to the questionnaire through the MOH unique ID number, which was also used on the questionnaire. In addition to free HIV counseling and testing, the counselor referred those who required services to the health facility. He/she also provided one-on-one education and counseling regarding risk reduction and prevention of HIV infection. He/she then encouraged respondents to repeat the test at a nearby HCT center after 3 months. All respondents (FSWs and LDTDs) who tested positive for HIV received information on available services and were referred to the health facility for appropriate care. 3.3.1 HIV Testing Clinical procedures for blood testing were carried out according to the national algorithm for rapid testing using Determine and Immunocomb kits. Trained laboratory technicians collected a single 18 biological specimen of 10 ml venous blood from each respondent using serum separation tubes labeled with the participant’s survey ID number and date of collection. The Determine test was then conducted immediately and results completed within 15 minutes. If the test result was negative, the participant received his/her result and post-test counseling. The remaining blood sample was then disposed of according to the national standard procedures. The positive and indeterminate results were tested in the health facility using the confirmatory Immunocomb test. After the confirmatory test, any discordant results were taken to the reference laboratory for further testing using the Western Blot technique. All samples were stored appropriately. 3.3.2 Syphilis Testing One drop of blood was used for conducting a rapid syphilis test in the field using the Determine syphilis kit. Results from the syphilis test were recorded on the paper form questionnaire. The blood sample for those who tested positive were used for confirmation using the Treponema pallidum Hem agglutination Assay. Those testing positive were also referred for treatment. The respondent was allowed to make an independent decision on whether to go to the health facility after counseling and receiving all relevant information. All data, including hard copies of informed consent forms and HIV/syphilis test results, were kept confidential and stored in a secure place under lock and key, accessible only to the key survey staff. 19 4 DATA MANAGEMENT AND ANALYSIS 4.1 Field Supervision For the IBBSS survey, the field coordinator provided supervision and oversight by reviewing all completed questionnaires and test results before they were uploaded to the server. During survey data collection, each site team consisting of interviewers, counselors and laboratory technicians had a supervisor verify all entries. The supervisor reviewed each completed questionnaire on the tablet and the HTC register immediately after the interview and testing session ended and identified mistakes for correction before the respondent left the site. Identified errors were discussed with the enumerators to avoid repetitions during succeeding interview sessions. A team composed of a field coordinator, the ROADS II M&E Officer and an ES/M&E officer in charge reviewed each completed questionnaire. The M&E officers were also involved in the supervision. After review and verification, the team provided guidance and oversight to supervisors where necessary. They also supervised the sampling techniques to ensure that the teams adhered to the methods described for each group and that all ethical issues were addressed. 4.2 Data Cleaning For the interviews, electronic versions of the final and approved questionnaires were uploaded into the electronic data collection device using a free and open-source data kit (ODK). HIV and syphilis test results obtained through specimen testing were entered in the electronic database and linked to the behavioral questionnaire using a unique ID code. To ensure the quality of data entry, a template for the questionnaire was designed with preprogrammed consistency checks for cross-checking answers, including skips and eligibility criteria. Before uploading the collected data to the server, a team composed of a field coordinator from MOH, a data manager, supervisors and a ROADS M&E officer proceeded on the data-cleaning exercise by reviewing each RA’s entries for consistency and completeness. A simple frequencies and cross-tabulation test was run to identify inconsistencies and/or detect missing values. Verification included:  Cross-checking all corresponding skips to the questionnaire  Tallying the supervisor’s log of blood samples collected to ensure that recorded numbers of samples matched the results recorded in the tablets and that the data were uploaded in formats compatible with the pre-approved analysis plan  Verifying that the questionnaire completion responses from the interviewers in the tablets matched those recorded in the supervisors log  Consistency checks involving cross-checking answers to related questions 4.3 Data Analysis Data were analyzed using the SPSS statistical analysis software. Simple frequencies were generated to initially recheck or revalidate the quality, and a range of relevant consistency checks were carried out. All the key behavioral and biological indicators (mostly in percentages or proportions) were calculated using appropriate numerators and denominators. Bi-variate analysis was conducted using cross-tabulations. HIV and syphilis results were stratified by socio-demographic characteristics and sexual behavior of survey populations. 20 Geographic coordinates of hotspots were collected using GPS-enabled electronic devices. Using ArcGIS software, maps were generated of hotspots and health services and juxtaposed with HIV prevalence and sexual behaviors. 21 5 LIMITATIONS OF THE SURVEY There was a possibility of the same respondent being interviewed more than once due to the associated mobility of the target groups. In this IBBSS, the participants (FSWs and LDTDs) were mobile. However, efforts were made to avoid such occurrences. Participants were asked before taking part in the survey if another survey team had interviewed them within the last month. Anyone interviewed in the last month was not re-interviewed. The survey did not cover all the potentially high-risk and vulnerable groups in the country. As such, the results from this survey are not necessarily representative of the key populations in Djibouti. The focus of this survey was on heterosexual activity. Same-sex relations are not explicitly included in this survey. The behavioral component of this survey exclusively relied on respondents’ self-reports. This data collection method often has limitations that are attributed to the tendency for people to under- report socially unacceptable attitudes and behaviors (e.g., risky sexual behavior) and to over-report socially acceptable behaviors (desirability bias). This survey was not designed to explain causality. One limitation of such cross-sectional design is that it is difficult to explain causality in both observed behavior and status. At the time of the survey, FSWs in PK12 were not available for recruitment and therefore only a small number of participants (12) were recruited. Most FSWs in this site are illegal immigrants (from Ethiopia and Somali) who feared being exposed to arrest if they participated in the survey. Data collection was extended by 5 days after the end of data collection in the other quartiers, but the anticipated number (31 and 125) was still not achieved. 22 6 RESULTS 6.1 Refusal to Take the HIV Test Of the 363 FSWs recruited for this survey, 361 (99%) agreed to participate in both behavioral interviews and the HIV/syphilis testing exercises. On the other hand, of 526 LDTDs selected, 23 (4%) refused to participate in the biological components of the survey. However, all 526 agreed to be interviewed. 6.2 Female Sex Workers 6.2.1 Socio-Demographic Characteristics of FSWs 6.2.1.1 Distribution of FSWs Based on the Area of Operation This survey recruited 363 FSWs in four quartiers in Djibouti Ville and 526 LDTDs (drivers and assistants) identified and selected from PK12 area. Table 4 shows the socio-demographic characteristics of the FSWs based on the location where they operated. Close to half (48.5%) of all FSWs sampled were located in the Quartier 2 area and more than 33% were in Arhiba. However, it was not possible to establish the reason why the FSWs preferred these two areas. The survey also revealed that a majority (52.5%) of the FSWs operated from service/bar and lodge, while 22% operated from homes and the remaining 25% operated in the streets, brothels or Chicha beits. This survey therefore identifies bars and residential homes as important contact points for FSWs and their clients. Such facilities should be given priority in future campaigns for safe sex. Table 4. Socio-Demographic Characteristics of the FSWs Location PK12 Quartier 2 Quartier 4 Arhiba Total Background characteristics % % % % N % Age group 18–24 33.3 38.1 37.0 42.1 142 39.1 25–29 58.3 26.7 37.0 25.6 105 28.9 30–34 8.3 13.6 14.8 20.7 58 16.0 35–39 0 11.9 5.6 7.4 33 9.1 40–44 0 9.7 5.6 4.2 25 6.9 Mean age (SD) 363 27.1 (6.7) Literacy Cannot read and write 41.7 75.6 85.2 79.3 280 77.1 Can read only 8.3 2.3 5.6 0.8 9 2.5 Can read and write 50.0 22.2 9.3 19.8 74 20.4 Based on where FSWs operate Brothel 0 5.7 0 20 34 9.4 Service bar/lodge 58.3 67.6 55.6 28.3 190 52.5 Chicha beits and other 8.3 1.1 0 24.2 32 8.8 Home 33.3 17.0 31.5 25.8 82 22.7 Street 0 8.5 13.0 1.7 24 6.6 Marital status Unmarried and living alone 41.7 63.6 74.1 19.8 181 49.9 Unmarried and living with partner 0 2.8 1.9 36.4 50 13.8 Married 16.7 0.6 1.9 2.5 7 1.9 Divorced/separated 33.3 28.5 18.4 34.7 106 29.2 Widowed and living alone 8.3 4.5 3.7 6.6 19 5.2 Total (SD) 12 (3.3) 176 (48.5) 54 (14.9) 121 (33.3) 363 100 23 6.2.1.2 Distribution of FSWs Based on Age The survey reveals that a majority (68%) of the FSWs were under the age of 30 and the mean age in this population was 27.1; most of the recruited FSWs were young, regardless of the area and facility from where they operated. Interestingly, only 7% of the FSWs were over the age of 40, probably indicating that FSWs retire only after a couple of years as SWs or that the older women migrate to other areas not covered in this survey. These assumptions remain to be ascertained in future studies. However, it is important to note that this survey focused only on SWs over the age of 18 and the results do not rule out the possibility that underage girls are involved in commercial sex. This survey did not cover FSWs who use other means of reaching their clients, such as via phone and social media. 6.2.1.3 Literacy among FSWs Significantly low levels of literacy were recorded among the FSWs. The illiteracy (inability to read and write) levels ranged between 41.7% and 85.2%, depending on the survey site, indicating that most FSWs were not properly educated. However, 50% of FSWs at PK12 could read and write. Thus, this survey suggests that a majority of the FSWs sampled may not be able to read and/or comprehend written information on critical such issues as HIV protection strategies, campaigns against HIV/AIDS and where to access support for sexual health, including condoms and contraceptives. 6.2.1.4 Marital Status among FSWs A majority (64%) of the recruited FSWs had never been married, including a larger group (49.9%) who were living alone and a second group (13.8%) who were unmarried but living with a partner. A significant percentage (29%) of FSWs were either divorced or separated. Only seven of the FSWs were married. However, it was not possible to establish if the spouses of the married FSWs were aware of their wives’ engagement in commercial sex. These data therefore reveal an important finding: that some FSWs maintain stable relationships despite their occupation. Such relations are likely to be highly risky, and the spouse may remain significantly predisposed to infection with HIV or STI. 6.2.2 Period of Work as an FSW The survey revealed that a high percentage (64%) of FSWs at Arhiba had worked for less than 3 years as SWs, while Quartier 4 had the highest proportion of FSWs who had worked for more than 10 years. The survey also revealed that more than 50% of all FSWs working in brothels, from their homes or from other public places such as bars and Chicha beits were relatively new to their profession, having worked for less than 3 years. FSWs in brothels were more likely to have worked there for more than 3 years than who operated in other facilities (OR: 5.79, CI: 3.41–9.83). The cross- tabulations of all statistical tests in this survey are presented in Appendix 1. 6.2.3 Mobility among FSWs Of the 363 females recruited in this survey, 69% lived in the same city or locale in which they operated as SWs. Another 21% were migrants from other regions of Djibouti, while 3% were from Ethiopia. Among the 363 FSWs recruited, only 100 (28%) agreed to reveal the identity of their native homes or country of origin. At least 64% of these 100 FSWs were from Djibouti, 29% were from Ethiopia and 7% were from Somalia/Somaliland. At least 29% of the 363 FSWs return home more than once per year. 24 6.2.4 Sex History among FSWs The survey shows that of the 363 FSWs recruited, 252 (69%) had their first sexual encounter when they were under 18 years old, but 72% started engaging in commercial sex after attaining the age of 18. Regardless of their category (where they operate), a majority of SWs (59%–89%) had a sex debut under the age of 18, but between 57% and 82% in different categories started engaging in transactional sex after the age of 18. The survey also reveals that those who had an early sex debut were more likely to start engaging in commercial sex at an earlier age than those who delayed their debut (OR: 6.61, CI: 4.77–9.16). 6.2.5 Transactional Sex among FSWs The current survey did not determine any statistical differences in type and number of sex partnersin Quartier 2 and Arhiba (the two sites with the highest concentration of FSWs) and those in the other regions (OR: 1.5, CI: 0.913–2.4). Our analysis did not identify a statistical difference between the number of sex partners among FSWs operating in defined facilities, such as brothels and homes, and those operating in public places, such as bars, restaurants and Chicha beits (OR: 0.66; CI: 0.42–1.04). 6.2.6 Sexual Partners among FSWs Majority of FSWs were found to engage with both regular sex workers and with occasional partners in which case least (63%) of the 363 participants had regular clients while 82% of these 363 also had occasional clients. High percentages (ranging between 50% and 56%) of FSWs working in all survey sites also had regular partners. A similar high percentage (ranging between 39% and 88%) of those working in different facilities, such as brothels, homes and public places, also had regular sex partners. The percentages of FSWs who had occasional clients in the four survey sites were also above 65%. Similar high percentages (75%–95%) of those operating in brothels, homes and public places had occasional partners. As expected, statistical tests showed that a higher proportion of FSWs were likely to have occasional clients than regular clients (OR: 3.2, CI: 2.3–4.5). 6.2.7 Family Planning 6.2.7.1 Use of Contraceptives among FSWs The methods and types of contraceptives that women use are important in HIV/STI prevention because not all contraceptives protect the user from infection while protecting her from unwanted pregnancies. Condoms are the only contraceptives that have this protective against both. There is a possibility that women who are not conversant with critical issues related to HIV and STI transmission may be keen to avoid pregnancy and therefore engage in risky unprotected sex if they are using other contraceptives, such as daily hormonal pills or injections that do not offer any protection against HIV and STIs. Although the main focus of this survey was on condom use, we also interrogated the use of other of contraceptives among the FSWs. Table 5 shows FP and contraceptive methods used by the participating FSWs. The current survey revealed that more than 60% of FSWs from the four survey sites use contraceptives. A total of 39% of the FSWs reported having previously conceived against their wishes, and 72% were on contraceptives at the time of recruitment. The condom was the most preferred contraceptive method (50%), followed by the daily hormonal pills (22%) and injectables (17%). Between 4% and 6% of those who used contraceptives employed a combination of at least two contraceptive methods. At least 38% of the FSWs used a combination of two FP methods. 25 Table 5. Family Planning and Contraceptive Use among FSWs by Socio-Demographic Characteristics Family planning (%) History of Contraceptive method unwanted On Background characteristics pregnancy contraceptives Daily pill Injectable Condom Age group 18–24 27.5 67.6 21.8 13.4 46.5 25–29 44.8 76.2 21.9 18.1 56.2 30–34 44.8 71.9 24.1 27.6 43.1 35–39 54.5 63.6 21.2 18.2 39.4 40+ 44.0 72.0 16.0 12.0 68.0 Location PK12 58.3 66.7 33.3 33.3 25.0 Quartier 2 42.6 65.3 11.4 6.8 57.4 Quartier 4 37.0 81.5 9.3 13.0 77.8 Arhiba 32.2 76.5 41.3 33.1 28.1 Based on where FSWs operate Brothel 32.4 57.6 29.4 17.6 14.7 Service bar/lodge 38.4 66.3 13.2 10.5 53.2 Chicha beits and other 40.6 87.1 53.1 34.4 43.8 Home 46.3 82.9 29.3 30.5 52.4 Street 25.0 75.0 12.5 4.2 70.8 Marital status Unmarried and living alone 37.6 71.7 12.2 12.7 60.2 Unmarried and living with partner 20.0 84.0 44.0 13.8 48.0 Married 57.1 57.1 42.9 1.9 0.0 Divorced/separated/widowed 50.0 60.0 30.0 20.0 40.0 All (%, n = 363) 141 (39.0) 256 (71.7) 79 (21.8) 63 (17.4) 180 (49.7) 6.2.8 Condom Use between FSWs and Their Clients Table 6 shows the pattern of condom use among FSWs and their occasional and regular partners in the last month prior to recruitment. Condom use with occasional clients was estimated at 82% (244 out of 297 who responded to the question) and more than 71% of all FSWs indicated that they consistently used condoms. Condom use with regular clients was 77% (238 out of 309 who responded to the question), while it was at 71% with the regular clients. These data indicate that between 29% and 40% of all sex encounters with different category of clients were unprotected. Condom use among FSWs working in brothels was only at 47%, while that of those working elsewhere was at 57%. Furthermore, only 38% of SWs in brothels used condoms regularly, compared to more than 48% of those working elsewhere. Only 50% of FSWs in Arhiba used condoms, compared to more than 66% of FSWs in all other survey sites. The number of FSWs in Arhiba who reported consistently using condoms with regular clients was also low (28%) compared to more than 88% in all other sites. These data suggest that, in general, condom use is low in Arhiba and in brothels. These data further support our hypothesis that FSWs working in brothels are likely to engage in risky sex because they establish a stable client base (regular customers) with whom they develop mutual trust. However, it is not clear why use of condoms in Arhiba was lower than use in other sites. More than 48% of FSWs were unable to use condoms because their clients refused to wear them, while 21% indicated that they would wish to have protected sex but that condoms were not available. Another 5.8% indicated that they did not use a condom because they trusted the 26 clients. Only a small percentage (7%) of those who did not use condoms cited affordability as the reason for not using condoms. Table 6. Condom Use among FSWs with Occasional and Regular Clients in the Last Sexual Intercourse Frequency of condom use with Frequency of used condoms with Condom occasional client Condom regular client use Most of use with Most of Background occasional Every the Some regular Every the Some characteristic clients time time times Never client time time times Never Age group 18–24 83.8 71.2 9.9 7.2 11.7 56.8 72.4 6.0 12.1 9.5 25–29 82.4 72.5 9.9 8.8 8.8 73.9 71.6 8.4 14.7 5.3 30–34 80.4 58.7 4.3 17.4 19.6 44.4 60.9 4.3 21.7 13.0 35–39 75.9 82.8 0 6.9 10.3 40.0 77.8 3.7 11.1 7.4 40+ 85.0 70.0 15.0 0 15.0 66.0 75.0 16.6 16.6 8.3 Location PK12 100.0 62.5 37.5 0 0 66.7 88.9 11.1 0 0 Quartier 2 95.7 90.0 7.1 2.1 0.7 66.7 91.1 4.8 3.4 0.7 Quartier 4 100.0 98.0 0 2.0 0 75.0 100.0 0 0 0 Arhiba 52.0 29.6 12.2 22.4 35.7 50.0 27.9 13.5 34.6 24.0 Based on where FSWs operate Brothel 59.3 33.3 11.1 18.5 37.0 45.5 37.5 9.4 28.1 25.0 Service bar/lodge 89.0 81.2 8.4 4.5 5.8 59.4 85.0 5.0 7.5 2.5 Chicha beits and other 70.8 50.0 8.3 8.3 33.3 77.8 48.1 22.2 14.8 14.8 Home-based 79.7 69.6 5.8 14.5 10.1 57.1 63.9 4.2 20.8 11.1 Street-based 81.8 72.7 9.1 9.1 9.1 66.7 75.0 12.5 6.3 6.3 244 210 25 26 36 238 219 22 41 25 All (%, n = 363) (82.2) (70.9) (8.1) (8.8) (12.2) (77.0) (71.3) (7.2) (13.4) (8.1) 6.2.9 Condom Breaks among FSWs Table 7 shows reported condom breakage among FSWs during sexual acts. In this group, FSWs reported a condom break 14% of the time in the month before recruitment %, while 9% reported a condom break during the last sexual intercourse. Table 7. Condom Breakage among FSWs Condom breakage (%) Condom break in the last Condom break in the last Intended to use condom Background characteristic one month sexual intercourse but did not Age group 18–24 13.4 9.9 13.4 25–29 12.4 6.7 18.1 30–34 17.2 8.6 10.3 35–39 18.2 15.2 12.1 40+ 11.8 5.9 16.0 Location PK12 8.3 8.3 8.3 Quartier 2 17.0 5.7 9.7 Quartier 4 5.6 9.3 9.3 Arhiba 13.2 13.2 24.0 Based on where FSWs operate Brothel 5.9 5.9 17.6 Service bar/lodge 14.7 7.9 10.0 Chicha beits and other 15.6 15.6 15.6 Home 12.2 8.5 18.3 Street 16.7 8.3 25.0 Total (%, n = 363) 50 (13.8) 32 (8.8) 52 (14.1) 27 6.2.10 STI/HIV/AIDS Knowledge, Attitudes and Behavior This section presents indicators of STI/HIV/AIDS knowledge, attitudes, and related behaviors for FSWs. It also seeks to establish respondents’ personal assessment of their risk of contracting STIs and HIV. 6.2.10.1 Knowledge of STI Symptoms and Infections among FSWs We sought to establish knowledge of STIs and health-seeking behavior among FSWs. Table 8 shows the knowledge of STIs and STI symptoms based on socio-demographic characteristics among FSWs. In general, 165 (46%) reported having heard of STIs. The most recognized STI symptoms in women by close to or more than 50% of the FSWs were: burning sensation (62%), lower abdominal pain (58%), foul-smelling vaginal discharge (51%), genital ulcers (49%) and itching around the genital area (59%). This group of FSWs was also aware of a number of STI symptoms in men, such as urethral discharge (44%), burning sensation during urination (52%), genital ulcers (41%) and swelling around the groin (33%). The proportion of FSWs who had heard of STI symptoms varied depending on the areas of operation. Only 18% of brothel-based FSWs reported having heard of STIs. On the other hand, 69% of those who operated from public places such as bars and Chicha beits had heard of STIs, followed by those who operate from home (60%). Table 8. Percentage of FSWs Who Ever Heard of an STI and Who Could Describe Any Symptom in Women and Men Aware of STIs Knowledge of STI symptoms in women Knowledge of STI symptoms in men Itching in genital area Swelling in groin area Foul-smelling vaginal Burning on urination Urethral discharge Genital ulcer/sore Genital ulcer/sore Lower abdominal Swelling on groin Burning/pain on Can’t retract discharge urination foreskin Background pain characteristics No % Age group 18–24 142 35.9 54.9 35.3 62.7 45.1 35.3 51.0 37.3 49.0 35.3 29.4 13.7 25–29 105 45.7 45.8 47.9 52.1 37.5 22.9 52.1 37.5 41.7 33.3 20.8 14.6 30–34 57 47.4 63.0 59.3 66.7 51.9 40.7 66.7 48.1 63.0 40.7 37.0 18.5 35–39 33 57.6 68.4 73.7 73.7 68.4 57.9 84.2 57.9 63.2 63.2 47.4 31.6 40+ 25 80.0 75.0 65.0 65.0 65.0 55.0 60.0 55.0 55.0 55.0 55.0 50.0 Location PK12 12 75.0 33.3 33.3 33.3 22.2 11.1 22.2 0.0 0.0 0.0 0.0 0.0 Quartier 2 176 42.0 47.3 55.4 55.4 55.4 35.1 54.1 45.9 51.4 51.4 33.8 33.8 Quartier 4 54 48.1 34.6 34.6 38.5 38.5 19.2 42.3 34.6 38.5 38.5 23.1 7.7 Arhiba 120 46.7 85.7 55.4 85.7 50.0 53.6 78.6 51.8 66.1 35.7 42.9 14.3 Based on where FSWs operate Brothel 34 17.6 83.3 50.0 83.3 33.3 16.7 83.3 16.7 66.7 50.0 0.0 16.7 Service bar/ 190 42.9 58 55.6 61.7 55.6 34.6 60.5 45.7 54.3 44.4 35.8 25.9 lodge Chicha beits 32 68.8 86.4 59.1 95.5 50.0 63.6 81.8 59.1 68.2 45.5 45.5 27.3 and other Home 82 59.8 40.8 42.9 44.9 38.8 32.7 44.9 36.7 36.7 32.7 26.5 12.2 Street 24 25.0 50.0 33.3 50.0 50.0 33.3 33.3 50.0 50.0 50.0 33.3 16.7 28 Aware of STIs Knowledge of STI symptoms in women Knowledge of STI symptoms in men Itching in genital area Swelling in groin area Foul-smelling vaginal Burning on urination Urethral discharge Genital ulcer/sore Genital ulcer/sore Lower abdominal Swelling on groin Burning/pain on Can’t retract discharge urination foreskin Background pain characteristics No % Marital status Unmarried and living 181 42.5 48.1 53.2 57.1 50.6 28.6 59.7 45.5 53.2 48.1 29.9 20.8 alone Unmarried and living 50 46.0 82.6 34.8 73.9 43.5 60.9 60.9 43.5 56.5 39.1 43.5 26.1 with partner Married 7 28.6 50.0 50.0 50.0 50.0 50.0 50.0 50.0 0.0 50.0 0.0 0.0 Divorced/ 105 51.4 63.0 53.7 63.0 48.1 38.9 59.3 44.4 50.0 33.3 35.2 22.2 separated Widowed and 19 47.4 44.4 55.6 66.7 55.6 44.4 44.4 22.2 44.4 33.3 33.3 11.1 living alone 164 94 84 101 80 61 96 72 84 68 54 35 Total 362 (45.3) (57.6) (50.9) (61.8) (49.1) (37.6) (58.8) (43.6) (51.5) (41.2) (33.3) (21.2) 6.2.10.2 Self-Reported Experience with STIs and STI Symptoms Table 9 shows data on self-reported cases of STIs and health-seeking behavior among the FSWs. Among 165 FSWs who reported having experienced STI-like symptoms in their lifetime, 19.0% had vaginal discharge, 16.3% had abdominal pain and 10.4% had genital ulcers. The proportion of FSWs who reported having had STI-like symptoms at Arhiba was at least twice as large as those who reported having STI-like symptoms in all other locations. We sought to establish health care-seeking behavior among the 121 FSWs who reported to have experienced STI-like symptoms in the last 1 year prior to recruitment. Among these 121 FSWs, 60% sought assistance from public health clinics and/or hospitals and 25% sought assistance from private pharmacies. Another 12% took home remedies, while 9% consulted private clinics and/or hospitals. About 38% of those aged over 40 and those aged 30–34 years (40.0%) preferred to buy medicine from private pharmacies. This survey revealed that a majority of FSWs continued to engage in sexual intercourse even when they had STI-like symptoms and that a majority did not reveal their illnesses to other people, including their sex partners or spouse and fellow FSWs. This is an important observation, considering that some STI-like symptoms reported included genital ulcers that are significantly associated with acquisition and transmission of HIV infections. 29 Table 9. Self-Reported Prevalence of STIs and Care/Assistance Sought by FSWs by Socio- Demographic Characteristics % reporting STI-like symptoms in the last 1 year Health-seeking behavior among those reporting STI-like symptoms Sought advise/medicine from a private clinic or Stopped having sex during the time when I had Sought advice/medicine from a non-allopathic Sought advice/medicine from a public clinic or Sought advise/medicine from a private Told my sexual partner about the STI Foul-smelling vaginal discharge Took medicine I had at home Lower abdominal pain Genital ulcer/sore Used condoms the symptoms pharmacy hospital hospital doctor Background characteristics Age group 18–24 17.6 18.3 7.8 90 0.0 20.0 0 20.0 10.0 20.0 0 25–29 12.5 14.9 6.3 55.6 0 22.2 0.0 22.2 11.1 11.1 18.2 30–34 33.3 26.9 19.2 41.7 23.1 38.4 7.7 7.7 0 0 7.7 35–39 26.3 26.3 21.0 80.0 0.0 40.0 20.0 0 25.0 20.0 0 40+ 10.0 15.0 5.0 25.0 25.0 .0 0.0 0 0.0 0 25.0 Location PK12 0.0 0 0.0 0 0.0 0.0 0.0 0.0 0.0 0 100.0 Quartier 2 17.6 10.8. 6.8 64.3 6.3 21.4 0.0 7.1 14.3 20.0 12.5 Quartier 4 11.5 11.5 0.0 0 0.0 25.0 0.0 0.0 0 0 0 Arhiba 26.8 29.4 21.8 65.2 13.0 30.4 9.1 4.3 4.3 4.5 4.3 Based on where FSWs operate Brothel 16.7 34.0 0 33.0 33.3 0 0.0 0 0 0.0 33.3 18.51 16.0 11.1 Service bar and lodge 52.9 5.0 22.2 6.8 11.1 11.1 16.7 15.0 2.8 Chicha beits and others 10.0 25.0 19.2 85.7 0 57.1 14.3 14.3 0 14.3 0 Home 20.4 12.5 8.2 58.36 15.4 25.0 8.3 8.3 0 0 0 Street 0 0 0 0 0.0 0.0 0.0 0.0 0.0 0 0.0 Marital status Unmarried and living alone 16..9 17.3 10.4 62.5 11.1 17.6 0 5.9 6.3 11.1 11.1 Unmarried and living with 19.0 10.0 9.1 80.0 0.0 12.2 0.0 40.0 20.0 20.0 0 partner Married 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0 Divorced/separated 20.4 20.4 11.1 46.7 6.3 33.3 13.3 13.3 6.6 13.3 12.5 Widowed 33.3 0.0 11.1 75.0 25.0 25.0 0.0 0.0 0 0 0 164 160 164 40 41 41 40 41 40 41 44 Total (19.0) (16.3) (10.4) (60.0) (9.1) (26.8) (5.1) (12.2) 5.1) (9.8) 9.1) 30 6.2.11 Prevalence of Syphilis among FSWs A total of 18 participants among the 361 FSWs who submitted blood samples were diagnosed with syphilis, giving an estimated prevalence of 5%. The rates were higher among the divorced (7.5%), those operating from brothels (11.8%), those aged over 40 years (20.0%) and those aged 30–34 years (8.6%) (see Table 10). Table 10. Prevalence of Syphilis among FSWs Percentage syphilis positive Number Age group 18–24 2.1 142 25–29 2.9 103 30–34 8.6 58 35–39 6.1 33 40+ 20.0 25 Location PK12 0.0 12 Quartier 2 6.3 175 Quartier 3 5.6 54 Arhiba 3.3 120 Based on where FSWs operate Brothel 11.8 34 Service bar/lodge 4.2 190 Chicha beits and other 3.1 32 Home 3.8 80 Street 8.3 24 Marital status Unmarried and living alone 5.0 180 Unmarried and living with partner 2.0 49 Married 0.0 7 Divorced/separated 7.5 106 Widowed 0.0 19 Total 5.0 361 6.2.11.1 Syphilis Prevalence Based on Age FSWs older than 30 years were more likely to test positive for syphilis than those under the aged less than 30 years (OR: 4.04, CI: 1.49–10.9). The syphilis prevalence among those older than 35 years was determined to be different from those observed among FSWs aged between 24–35 years (OR: 1.4, CI: 0.55–3.62). It is therefore possible that older FSWs who have engaged in commercial sex for longer than the younger females may have a have higher probability of acquiring syphilis infections. This explanation, however, assumes that a majority of FSWs start engaging in commercial sex at an early age and that older FSWs have been in this business for a longer period of time. 31 6.2.11.2 Syphilis Prevalence Based on Literacy Good literacy skills may allow a person to access important information on HIV and STI prevention through better comprehension of verbal and written information. However, the literacy level was not found as a significant influencing factor on the outcome of syphilis testing among FSWs (OR: 2.18, CI: 0.44–9.7), suggesting that all FSWs may be equally predisposed to these infections regardless of their literacy level. This is not surprising considering the fact that the majority of FSWs were not properly educated and that even those who had “higher” levels of literacy were not schooled for more than 10 years. 6.2.11.3 Syphilis Prevalence Based on Marital Status This survey showed that there was a relationship between marital status and HIV status among FSWs. However, being divorced, separated, or widowed was not found as begin significantly associated with the likelihood of FSWs testing positive for syphilis (OR: 1.4, CI: 0.22–1.56). 6.2.11.4 Syphilis Prevalence Based on Place of Operation for FSWs Our analysis did not identify significant differences in syphilis prevalence among FSWs based on their operation base (OR: 0. 44, CI: 0.17–1.15). These findings suggest that the risk of infection with syphilis may be equally spread across all survey sites from which the FSWs were recruited. 6.2.11.5 Syphilis Prevalence Based on Age at Sex Debut and Exposure to Transactional Sex Our analysis did not find a significant association between the age of sex debut of an FSW and the outcome of her syphilis status (OR: 1.35, CI: 0.51–3.59). Similarly, the age at which a female started engaging in commercial sex was not found to be related to her syphilis status (OR: 1.17, CI: 0.37– 3.69). The most logical explanation for these observations is that the risk of syphilis infection rises due to a combination of factors and that the age of sex debut alone may not be the most significant factor. 6.2.11.6 Syphilis Prevalence Based on Type and Number of Sex Partners We failed to determine any association between the number of partners an FSW had and the outcome of the syphilis test (OR: 0.92, CI: 0.05–16.71). The chances of testing positive for syphilis were found to be similar between those FSWs engaging in sex with fewer than five clients per week and those engaging with more than five clients (OR: 0.49, CI: 0.16–1.53). Those who had regular clients were found to be just as likely to be syphilis positive as those who had occasional clients (OR: 0.96, CI: 0.47–1.95). These results do not necessarily indicate that the number of sex partners and the volume of sex per week per FSW does not increase the chances of acquiring syphilis, but instead may point to the fact that the manner in which transactional sex is conducted (especially if condoms are not used) is more important in STI prevalence than how often the FSWs engage in sex. 6.2.11.7 Syphilis Prevalence Based on Length of Service The length of service as a SW was found to be significantly associated with the outcome of syphilis status (OR: 0.47, CI: 0.17–1.27). These results may be interpreted to mean that no matter how long an FSW works, the likelihood of contracting syphilis remains the same. 6.2.11.8 Syphilis Prevalence Based on Use of Contraceptives The chance of testing positive for syphilis among FSWs who were using contraceptives was not determined to be significantly different from that of those not using them (OR: 1.909, CI: 0.58–6.79). 32 These results suggest that as long as condom use remains low among FSWs, uses of other contraceptives that do not prevent the exchange of sexual fluids apparently have no significant influence on the prevalence of syphilis. 6.2.11.9 Syphilis Based on Exposure to Anal Sex Anal sex was identified as one of the most important risk factors for syphilis prevalence among FSWs (OR: 10.105, CI: 3.64–28.04). These results suggest that anal sex is a risk factor for syphilis, just as it is for HIV infection. 6.2.11.10 Syphilis Prevalence Based on Unprotected Sex and Condom Breaks FSWs who reported condom breaks were found to be more likely to test positive for syphilis than those who had unprotected sex (OR: 2.56, CI: 0.87–7.54). The most plausible explanation to these seemingly perplexing observations is that due to the low usage of condoms among FSWs, a condom break does not necessarily present a bigger risk for an infection than that presented by unprotected sex, which is rampant in this population. 6.2.11.11 Syphilis Prevalence Based on Alcohol FSWs who take alcohol on daily basis were found to be more likely to test positive for syphilis than those who had never taken alcohol (OR: 5.01, CI: 1.72–14.58). The reasons for these differences remain unclear and further studies should be conducted to investigate this phenomenon. 6.2.11.12 Syphilis Prevalence Based on Drugs and Sharing of Needles One of the most unexpected results in this survey was that the use of drugs and even sharing of needles were not determined to be significantly related to the outcome of a syphilis test (OR: 1.90, CI: 0.59–7.79). While these results are baffling, it is important to realize that a very small proportion of participants used drugs and only 6% of FSWs shared needles. Within this population, two were HIV positive. It is therefore possible that a larger sample size than what we used in this survey could reveal more conclusive data on the role of sharing needles in transmission of syphilis among FSWs. There is also a possibility that drug use is under-reported in this country due to cultural and religious reasons. 6.2.11.13 Syphilis Prevalence Based on GBV Our survey found that women who reported cases of GBV were more likely to test positive for syphilis than those who did not (OR: 0.219, CI: 0.03–1.68). The most likely explanation for these results is that some forms of GBV may be of a sexual nature and probably unprotected. This could significantly increase the exposure of the victims to infections, including syphilis. 6.2.11.14 Syphilis Prevalence Based on Knowledge of HIV/STIs and Self-Assessment of Possible Risk to Infection Our survey determined that FSWs who believed that they were at risk of HIV infection/STIs and those who did not know if they were at risk had the same chances of testing positive for syphilis (OR: 0.46, CI: 0.14–1.5). This observation may be due to a combination of several factors, including low literacy levels and poor knowledge of HIV and STI issues among the FSWs. 33 6.2.11.15 Self-Assessment of Risk to HIV and Knowledge of HIV/AIDS Support Programs among FSWs Knowledge of HIV status helps HIV-negative individuals make specific decisions to reduce risk and practice safer sex. For those who have HIV, knowledge of their status allows them to take action to protect their sexual partners, to access treatment and to plan for the future. To assess the awareness and coverage of HIV testing services, respondents were asked whether they had ever been tested for HIV. If they responded affirmatively, they were asked whether they had received the results of their last test and when the testing occurred. Table 11 shows FSWs’ perception of self-risk of HIV infection, HIV testing status and knowledge of ART by socio- demographic characteristics. At least 40% of FSWs were aware that there are drugs that can help manage HIV/AIDS among infected people. At least 25% of FSWs had heard of ART and 15% knew of someone who was taking ART drugs. Self-assessment of possible modes of acquiring and transmitting HIV and STIs and knowledge of risk factors play important roles in prevention of infection transmission. About 40% of the FSWs felt at risk of HIV infection. More than half (54%) reported having been tested for HIV previously, and 96% of those tested reported that the tests were voluntary and that they received their results upon testing. Surprisingly, 56% of drug users and half (50%) of those who engaged in anal sex did not know that they were exposed to HIV and STIs. Close to 60% of FSWs who had sex with more than five clients per week, those who worked daily and those who had sex without a condom did not necessarily consider themselves at risk of HIV infection. A similar proportion of FSWs who suffered GBV did not consider themselves at risk. A worrying observation was that 26 of the 47 HIV sero- positive cases (55%) and 5 of the 18 syphilis-positive cases (28%) reported in this survey were detected among FSWs who did not know they were at risk. 6.2.12 HIV Prevalence among FSWs Figure 1 shows HIV prevalence by socio-demographic characteristics. Overall, the HIV prevalence among FSWs was estimated at 13%. The prevalence was lowest for those aged 18–24 years (5%). HIV prevalence among those over the age of 40 was as high as 36%. We analyzed for HIV prevalence based on selected characteristics as presented in the sections below. Figure 1. HIV Prevalence among FSWs by Age 34 6.2.12.1 HIV Prevalence Based on Age This survey showed strong association between HIV prevalence among FSWs with increasing age. Those aged 30 years and older were more likely to be HIV positive than those aged less 30 years (OR: 6.56, CI:3.2–13.5). These data suggest that the risk of acquiring HIV infection increases with increasing age. These findings confirm those showing that FSWs older than 35 years were significantly more likely to test positive for HIV. 6.2.12.2 HIV Prevalence Based on Literacy Considering that literacy plays an important role in helping individuals gain meaningful knowledge, access information that enables them to make informed decisions and use their knowledge to discern health risks, it is possible that low literacy levels may compromise the chances of accessing such information among the participants. We analyzed for HIV prevalence based on the level of education and found that 34 of the 47 HIV-positive cases (72%) were among the illiterate group of FSWs, but these proportions were not determined to be statistically different from those among the “literate” groups (OR: 1.39, CI:0.64–2.98). 6.2.12.3 HIV Prevalence Based on Marital Status It was observed that 26 of the 47 HIV-positive cases (55%) reported among FSWs were detected among women who were divorced, separated or widowed. Further analysis revealed that divorced/separate/widowed FSWs were more likely to test positive than those who were unmarried (OR: 2.36, CI: 1.18–4.77), but not than those who were married (OR: 0.36, CI: 0.04–3.4). These results are in agreement with others that show that HIV cases were relatively more common among the same group of divorced/separated/widowed FSWs. As mentioned before, we speculate that some of the widowed FSWs may have lost their partners to HIV, hence the higher prevalence of HIV in this group. Future studies should ascertain whether these assumptions are true. 6.2.12.4 HIV Prevalence Based on Base of Operation for FSWs Close to half of the 47 HIV-positive cases (23, 49%) among FSWs were from the group that operated in public places, such as Chicha beits and bars. However, these proportions were not significantly different from those recorded among FSWs operating in brothels and other types of accommodations (OR: 0.61, CI: 0.33–1.13). Although 38 (80%) of HIV-positive cases were recorded among FSWs operating in Arhiba and Quartier 2, the proportions of HIV-positive individuals in these two locations were found to be significantly different from those observed in other survey sites (OR: 3.03, CI: 0.73–6.09). These observations suggest that the chances of acquiring HIV infections may be equally distributed across all the sites. 6.2.12.5 HIV Prevalence Based on Age at Sex Debut and Age at Exposure to Transactional Sex An early sex debut can be a risk factor for acquiring STIs. We sought to categorize HIV prevalence based on the age at which participants had their first sexual encounters and the age at which they started engaging in transactional sex. Of the 47 cases HIV, 25 (53%) were among FSWs who had a sex debut under the age of 18, and the remaining 22 were among those who had a delayed debut. However, HIV prevalence in these two groups were not statistically different (OR: 0.96, CI: 0.52– 1.78). 35 6.2.12.6 HIV Prevalence Based on Type and Number of Sex Partners The number and type of sex partners can be a compounding factor in HIV acquisition and transmission. This survey was not able to demonstrate a significant association between the number of clients that FSWs engaged with and HIV test outcomes (OR: 0.65, CI: 0.31–1.4). This is probably because more than 60% of FSWs had both occasional and regular clients and condom use was not consistent. It is important to note that some FSWs revealed that they had maintained stable sexual relationships with regular male clients for a period of up to 10 years. It was also noted that unprotected sex with regular clients was common. Such sexual behavior is likely to increase the possibility that, although the type of client an FSW has may not be related the outcome of her HIV test ,engaging in unprotected sex is likely to be a more risky behavior in HIV infection than the type and number of clients the FSW has. 6.2.12.7 HIV Prevalence Based on Length of Service The current survey revealed that 25 (53%) of all HIV-positive cases were detected among FSWs who had been in this profession for a period of 3 years or more. Statistical analysis revealed that FSWs in this category were more likely to test positive for HIV than those who had worked for less than 3 years (OR: 2.25, CI: 1.15–4.42). It is therefore possible that working as an FSW for a long time increases the probability of contracting HIV. 6.2.12.8 HIV Prevalence Based on Use of Contraceptives This survey found that condoms were the most preferred form of contraceptive, but that their use among FSWs was only at around 50%. A majority of women were found to use a combination of methods to prevent pregnancy. A high percentage, 34 (72%) of the 47 HIV-positive FSW cases, was detected among FSWs who were using contraceptives. The HIV prevalence was similar for those using a contraceptive method and those not using (OR=1.15, CI:0.57-2.45). These results suggests that, in the absence of high condom use among FSWs, the use of other contraceptives may not influence the HIV test outcomes. There is therefore a need to sensitize this population on the need to use condoms as safer contraceptives in every sexual encounter, because condoms provide dual protection against infections and pregnancy. 6.2.12.9 HIV Prevalence Based on Exposure to Anal Sex Unprotected anal sex is recognized as an important risk factor for HIV acquisition among males and females. We investigated if there was a association between engaging in anal sex and acquisition of HIV among FSWs and found that only 4 FSW ever had anal sex and were HIV negative. This is surprising considering that condom use was significantly low. It is important to consider that this survey also revealed that FSWs received more requests for anal sex but the majority did not consent for the same. This calls for improved education strategies and campaigns to sensitize FSWs and their clients on the need to use condoms in all sex encounters, especially anal sex, as HIV prevalence will continue to rise and probably result in similar trends to those observed among homosexual males who have been identified as particularly vulnerable to HIV. 6.2.12.10 HIV Prevalence Based on Unprotected Sex and Condom Breaks This survey showed that there was a significant association between condom use and testing HIV positive among the FSWs (OR: 3.07, CI: 1.31–7.17), and those reporting experience of condom breaking were more likely to be HIV positive (OR: 1.59, CI: 0.66–3.77). We recommend that more 36 robust case-controls experiments should be designed in future to investigate this phenomenon further. In addition, since unprotected sex is only about 50% among the FSWs, it is possible that the effect of a condom break is masked by the risk associated with not using a condom at all. 6.2.12.11 HIV Prevalence Based on Use of Alcohol Alcohol is reported to distort judgment, and an intoxicated person is more likely to engage in risky behavior than the one who is sober. As mentioned in preceding sections, alcohol use was not found to be a serious problem in both target groups (FSWs and LDTDs), possibly because a majority of people in Djibouti are Muslim and alcohol use may not be popular in such settings. Indeed, statistical analysis failed to draw any associations between consumption of alcohol and HIV status among FSWs (OR: 0.31, CI: 0.16–0.94). These results were similar to those on syphilis that showed no association between alcohol intake frequency and being syphilis positive (OR: 0.61, CI: 0.16–2.12). The reasons for these differences remain unclear.) 6.2.12.12 HIV Prevalence Based on Drug Use and Sharing of Needles Using drugs and sharing needles have been identified as important factors for HIV acquisition. Sharing contaminated needles is one of the most risky behaviors associated with HIV acquisition, partly because a high viral dose may directly reach the blood stream without being subjected to screening by first lines of immunological defenses. This survey found that among the 120 FSWs who used drugs, 29 (24%) used injectable drugs and shared needles. The use of drugs was not significantly associated with the HIV prevalence among FSWs (OR: 4.88, CI: 0.25–94.84). Similarly, this survey did not show statistical differences on HIV prevalence among drug users who shared needles and those who did not (OR: 0.84, CI: 0.43–1.63). Considering that of the 29 FSWs who shared needles, 2 tested positive for HIV, there is a chance that a robust sample size in a future survey will link sharing needles and HIV infections in vulnerable populations in Djibouti. We also suspect that the use of drugs, especially injecting drugs and sharing needles, may be under-reported among the participants due to the possibility of stigma associated with use of drugs. 6.2.12.13 HIV Prevalence Based on GBV GBV makes women venerable to HIV because they may be subjected to unprotected sex, which may cause physical trauma to the genitals and predispose females to acquire an infection. It is expected that in most cases FSWs may not be in a position to negotiate for safe sex, especially where GBV is involved. Some cases of GBV may also involve sodomy, and unprotected anal sex is significantly associated with HIV infections. This survey found that HIV prevalence among the 74 women subjected to GBV was 10%, but those who reported being a victim of GBV were not necessarily found to have a higher probability to test positive for HIV than those who did not (OR: 0.65, CI: 0.27– 1.52). These data did not demonstrate GBV alone as a risk factor in HIV acquisition, but a combination of compounding factors, including whether or not the GBV involves unprotected sex, is more important in HIV infections among FSWs in Djibouti. 6.2.12.14 HIV Prevalence Based on Knowledge of HIV and Self-Assessment of Risk to HIV Infection This survey revealed that 31 of the 47 HIV-positive cases (66%) among FSWs were among people who knew about HIV/AIDS. Statistical analysis revealed that FSWs who did not have sufficient knowledge of HIV were more likely to test positive for HIV than those who were well versed with critical issues (OR: 0.37, CI: 0.19–0.075). However, self-assessment of risk to HIV (participant 37 perceiving that she was at risk, was not at risk or not knowing whether she was at risk or not) those who think they are at risk were more likely to be HIV positive than those not perceiving themselves as at risk (OR: 2.22, CI: 0.99–5.05). 6.2.12.15 HIV Prevalence Based on Previous HIV Testing This survey revealed that the chances of testing positive for HIV were similar between those who had been tested in the last 12 months and those who had not tested.(OR: 0.83, CI: 0.07–15.15). 6.2.13 Gender-Based Violence In recent years, there has been increasing concern about violence against women in general, and about domestic violence in particular, in both developed and developing countries. Not only has domestic violence against women been acknowledged as a violation of basic human rights, but also an increasing amount of research highlights the health burdens, generational effects and demographic consequences of such violence (United Nations General Assembly 1991; Heise et al. 1994; Heise et al. 1999; Jejeebhoy 1998). Women also bear the health and psychological burdens of most GBV. Victims of domestic violence are abused inside what should be a secure environment— their own homes. The collected data revealed that 74 cases of GBV were recorded, but only 50% of these cases were reported. Only one of these cases was reported to a nongovernmental organization (NGO), and only 19% and 18% of victims were willing to report such cases to the police or a friend, respectively (see Table 11). Of these GBV cases, 28% were of a sexual nature. Although this survey did not establish whether or not some of these cases involved sodomy, 75% of the FSWs who reported having had forced sex also reported to have had anal sex. A majority of the perpetrators of these acts were clients. Relatives were involved in 4% of these cases, while friends were reported to have committed 11% of these cases. As mentioned in the preceding sections, our analysis was unable to show any association between GBV and HIV status (OR: 0.65, CI: 0.27–1.52), but those subjected to GBV were apparently more likely to test positive for syphilis (OR: 0.219, CI: 0.028–1.68). These data suggest that, although both of these infections are sexually transmitted, their transmission dynamics may be different. 38 Table 11. Experience on GBV GBV cases N = 74 % Reported 37 50 Reported to: NGO 1 1 Police 14 19 Friend (FSW) 13 18 Friend (FSW) + police 9 12 GBV cases involving forced sex 28 38 Among those who reported of having had anal sex (N = 28) 21 75 Perpetrators of GBV N = 28 % Forced by client 11 39 Forced by main (regular) client 6 21 Forced by a friend 3 11 Forced by relative 1 4 6.2.14 Mapping of FSWs Hotspots in Relation to Facilities and Resources in the Survey Area Geographic information system (GIS) mapping of FSW hotspots and other facilities, including heath care centers, administrative facilities, NGOs and other resource centers, revealed that the mapped points were located in areas with good road networks and access to these facilities should be easy. These maps are available and can be provided on request, but we did not include them in this report based on ethical considerations and the safety of the SWs. In general, the distribution of resource centers in Arhiba was poor, with centers mapped in the region located quite a distance from each other. Similarly, although a majority of SWs were mapped around the quartiers, the distribution of administrative support, NGOs and health care facilities was poor in these locations. The resource centers at PK12 were also located far outside the range zone of concentration of FSW hotspots. Only one major health care facility was mapped around PK12, and none of these facilities were present at the PK12 satellite site (truck bay). 6.3 Long-Distance Truck Drivers 6.3.1 Socio-Demographic Characteristics of LDTDs Table 12 shows the distribution of socio-demographic characteristics of LDTDs. More than half of the truckers were younger than 30 years, with a mean age of 31 years, and only 16% were aged 40 years or over. 39 Table 12. Socio-Demographic Characteristics of Truckers Total Background characteristics N % Age group < 20 25 4.8 20–24 136 25.9 25–29 116 22.1 30–34 89 16.9 35–39 75 14.3 40+ 85 16.1 Mean age (SD) 526 31.0 (10.1) Literacy Cannot read and write 181 34.4 Can read only 11 2.1 Can read and write 334 63.5 Marital status Currently married 236 44.9 Divorced/Separated/Widowed 14 2.7 Never married 269 51.1 No response 7 1.3 Total 526 100 6.3.1.1 Habitation of LDTDs A majority of truckers, 493 (93%), slept in their trucks during the road trips, while only 28 (5%) and 8 (2%) slept either in a home or in a hotel, respectively during their field trip. It was however not established if those who slept in homes during the field trips lived with their wives, relatives or sex partners. 6.3.1.2 Literacy among LDTDs Literacy levels were higher among the truckers than among the FSWs, with 64% of the LDTDs being able to read and write This suggests that the LDTDs are likely to be more advantaged as far as accessing and understanding written information on critical issues on HIV and STIs. However, it is important to note that, although the LDTD population was significantly more literate than was the FSW population, less than 20% of the truckers had an education of more than 10 years of schooling. This observation further suggests that even though the LDTDs are more literate than the FSWs, they are equally likely to experience serious limitations in understanding complex issues regarding sexual health, prevention of HIV and STIs and linking various behavioral components to infections and transmission of sexually acquired infections. 6.3.1.3 Marital Status among LDTDs Almost 45% of all LDTDs were married at the time of recruitment; 51% had never been married and 3% compared to 29% among FSWs were either divorced/separated/widowed. . At least 94% of all the married truckers lived in a trucks during the road trips, and among unmarried truckers, 95% of 40 whom lived/slept in their trucks during road trips. However, considering that a majority of these married truckers are Ethiopians, 6.3.2 Period of Work as an LDTD At least 331 (63%) of all truckers had worked for less than 3 years, indicating that most LDTDs were still new in their jobs. Our analysis did not find a significant association between the length of time a trucker had worked and his ability to seek other types of accommodation, rather than sleep in his truck (OR: 0.23, CI: 0.028–1.88). It is therefore possible that factors other than familiarity with the truck stop and the period during which the trucker had worked made a majority of truckers prefer to sleep in their trucks. 6.3.3 Mobility among LDTDs At least 27% of the LDTDs indicated that they did not visit their home or native districts or families. These data further suggest that a majority of truckers live a life of physical alienation from their families and maybe friends, who could provide support. In contrast to the FSWs (a majority of whom prefer to operate in their native homes), a majority (more than 60%) of LDTDs were from Ethiopia, indicating that the truckers’ population was largely made up of foreigners. A large majority (93%) of these truckers were travelling to Addis Ababa and other towns in Ethiopia, while the rest were destined to the port of Djibouti. The LDTD group was identified as highly mobile, with 443 (84%) making a round trip to their destinations in 3 days. At least 15% of the truckers made between 26 and 30 round trips per month. These data suggest that a majority of the truckers are away from their homes for long periods of time, during which they live in sub-optimum conditions (their trucks). In addition, the numerous trips they make suggest that they are likely to be overworked. A combination of such factors is likely to force truckers to engage in risky behavior that may include unprotected sex. It is important to remember that a majority of these foreign truckers have a relatively low level of education and may also suffer from culture and language barriers. An interaction of such socio-cultural and economic challenges and low literacy levels may make LDTDs significantly vulnerable to HIV and STIs. This vulnerability is multiplied by the fact that the truckers solicit sex from FSWs who have even lower literacy levels. 6.3.4 Sex History among LDTDs Among the LDTDs recruited in this survey, 77% had their first sexual encounter when they were between the age of 16 and 20 years. Seven of the participants had not had any sexual encounter in their life. This survey revealed that those who had a sex debut at an age under 15 years were more likely to seek a paid sexual encounter than those who had a delayed debut (OR: 2.3, CI: 1.4–10). This survey also revealed that 77% first sexual encounters among the LDTDs involved non-paid females. A majority of truckers reported having had their sex debut with a friend (67%); others reported their sex debut with a co-worker (40%) or a stranger (33%). This survey also established that cases of incestuous first sexual relationships occur in about 15% among LDTDs. 6.3.5 Where Do LDTDs Pick Sex Partners? At least 10% of LDTDs indicated that they had paid for sex in the last 1 month. Brothels were the preferred sources of sex partners among the LDTDs. Of the LDTDs interviewed, 5% picked their partners during loading and reloading of their trucks, and a similar proportion picked FSWs when staying at home during the loading/reloading. Statistical analysis revealed that the chances of 41 picking a SW was equal between LDTDs living/sleeping in their trucks and those seeking accommodation in a hotel or a home (OR: 1.46, CI: 0.5–4.2). Considering that most truckers live in their trucks and most of them reported to have picked an FSW in the course of their work, this survey indirectly suggests that some truckers may take their partners to the trucks to perform sex acts, but this remains to be verified. Should this be the case, it is important for future studies to investigate if sex acts in a truck are performed in privacy and if there are cases of group sex or shared sex partners. It is also important to investigate whether the truckers, especially those who live in homes, perform sex acts in brothels and lodges or take their sex partners home or use the trucks for sex. 6.3.6 Condom Use among LDTDs This survey found that up to 10% of LDTDs did not use condoms with their sex partners. Although this proportion of LDTDs might appear low, not using condoms with FSWs is particularly risky because FSW population has been identified as an important seed population that fuels the emergence of new HIV infections in most countries. The 10% of LDTDs who did not use a condom could in turn infect other partners, including their spouses Table 13 shows condom use during sexual intercourse with paid and non-paid female partners by truckers. About one-fifth of the truckers reported having paid for sex, and 90% of these had used a condom in their last sexual activity prior to recruitment. Among those who used a condom, 87% indicated that they consistently used condoms with a paid sex partner. Another 25% of truckers indicated having had sex with a non-paid partner, and 76% of these had used a condom with their non-paid partners. Among those who used condoms with occasional partners, 74% consistently used condoms. These data show that, depending on the category of the sex partner, at least 13% to 26% of sexual encounters by LDTDs are not protected. Those not using condoms are likely to be particularly vulnerable to infections and may act as a seed population to other people, including their spouses and other sex partners. 42 Table 13. Condom Use with Paid and Non-Paid Female Sex Partners among LDTDs Condom use with paid female sex partner (%) per category Condom use with non-paid female sex partner (%) per category Frequency of condom use in last year Frequency of condom use in last year Sex Reported with paid partner with non-paid partner with a Condom use with Sex with a Condom use with a condom Background paid paid partner in the Most of non-paid non-paid partner in Most of break in last characteristic partner last 1 month Every time the time Sometimes partner the last 1 month Every time the time Sometimes 1 year Age group < 20 9.5 100.0 100.0 0.0 0.0 14.3 100.0 100.0 0.0 0.0 4.0 20–24 14.9 90.0 86.7 13.3 0.0 40.3 75.9 75.5 19.5 4.9 1.5 25–29 30.2 85.7 77.8 11.1 11.1 41.4 85.4 70.7 24.4 4.9 7.8 30–34 23.9 85.7 83.3 16.7 0.0 18.2 68.8 63.6 18.2 9.1 3.4 35–39 20.0 100.0 100.0 0.0 0.0 13.3 70.0 85.7 14.3 0.0 2.7 40–44 16.7 66.7 100.0 0.0 0.0 6.7 100.0 50.0 50.0 0.0 0.0 45+ 7.3 90.2 100.0 0.0 0.0 10.9 16.7 100.0 0.0 0.0 0.0 Marital status Currently married 13.6 86.4 100.0 0.0 0.0 8.9 47.6 70.0 10.0 10.0 2.1 Divorced/separated/ widowed 50.0 100.0 100.0 0.0 0.0 35.7 100.0 80.0 20.0 0.0 14.3 Never married 23.4 90.6 79.2 14.6 6.3 42.0 80.5 73.6 22.0 4.4 3.7 102 74 64 7 3 139 106 102 29 6 17 All (out of 526) (19.7) (90.2) (86.5) (9.5) (4.1) (26.8) (76.3) (73.6) (20.8) (4.7) (3.3) 43 6.3.6.1 Condom Breaks among LDTDs In general, condom breakage during sex was reported at lower frequencies among the LDTDs than among the FSWs. Overall, only 3.3% of the 526 LDTDs reported experiencing a condom break. The percentage of those reporting condom breakage was below 10% in all categories except among the divorced/separated/widowed, which was 14%. 6.3.7 Knowledge of STI Symptoms and Infections among LDTDs Table 14 shows the knowledge of STIs and STI symptoms by socio-demographic characteristics of LDTDs. Knowledge of these symptoms was relatively low, with only 6.5% reporting to have heard of STIs. FSWs were more likely to identify at least one STI symptom unassisted compared to LDTDs . At least 90% of the 30 LDTDs who knew about STIs could identify “a burning sensation during urination” as an STI symptom. Other symptoms correctly identified by these 30 LDTDs included urethral discharge (40%), swelling of groin (40%), inability to retract foreskin (30%) and genital ulcers (16.7%). No LDTD reported having had symptoms of STIs and therefore none were interviewed on STI-related health-seeking behavior Table 14. Percentage of LDTDs Who Ever Heard of an STI and Could Describe Any Symptom in Men Ever heard of diseases transmitted through sexual intercourse Can describe any Symptoms of STIs in men (%) Swelling in Genital groin Cannot Background Total Urethral ulcers (scrotal) Burning pain retract characteristics N Percentage N discharge or sores area on urination foreskin Age group < 20 22 4.5 1 0.0 0.0 0.0 100.0 0.0 20–24 116 9.5 11 36.4 18.2 36.4 90.9 0.0 25–29 103 4.9 5 60.0 40.0 60.0 60.0 20.0 30–34 79 7.6 6 50.0 0.0 33.3 100.0 50.0 35–39 73 1.4 1 0.0 0.0 0.0 100.0 100.0 40–44 26 3.8 1 0.0 0.0 100.0 100.0 100.0 45+ 51 9.8 5 40.0 20.0 40.0 100.0 60.0 Marital status Currently married 215 4.2 9 66.7 33.3 44.4 100.0 2.1 Divorced/separated/ 12 16.7 2 0.0 0.0 0.00 100.0 7.1 widowed Never married 237 8.0 19 31.6 10.5 42.1 84.2 1.1 Total 470 6.5 30 12 (40.0) 5 (16.7) 12 (40.0) 27 (90.0) 9 (30.0) * No response=6 6.3.8 Prevalence of Syphilis among LDTDs The syphilis prevalence rate among the 526 truckers who submitted blood samples was estimated at 1.6%. High values were recorded among those who were aged over the age of 45 years (5.6%), those between the ages of 30 and 34 (2.3%) and those between the ages of 35 and 40 years (1.4%). A high value was also observed among those who were married, namely, 2.6%. 44 Table 15. Prevalence of Syphilis among LDTDs Syphilis status Percentage syphilis positive Number p-value Age group < 20 0.0 24 20–24 0.8 133 25–29 0.9 107 30–34 2.3 86 35–39 1.4 69 40–44 0.0 30 45+ 5.6 54 Marital status Currently married 2.6 222 Divorced/separated/widowed 0.0 12 Never married 0.8 255 Total 1.6 503 0.244 6.3.8.1 Syphilis Prevalence Based on Age Of the eight syphilis cases detected among LDTDs, four (50%) were among those over the age of 35 years, and those age 35 year and higher were more likely to be syphilis positive than those below 35years but not statistically significant difference (OR: 2.32, CI: 0.48–11.2). These data suggest that age could be an important factor in the prevalence patterns of syphilis among the FSWs, but not among LDTDs. 6.3.8.2 Syphilis Prevalence Based on Literacy Similar to FSWs, the literacy level was not associated with the outcome of syphilis testing among LDTDs (OR: 0.92, CI: 0.22–3.87). This was similar to the no association between syphilis and literacy level among FSWs (OR: 0.40, CI: 0.06–1.09). It is important to consider that a majority of truckers were Ethiopians and there is a possibility that language and cultural barriers hinder their access to information. 6.3.8.3 Syphilis Prevalence Based on Marital Status Although being divorced was identified as a risk factor among FSWs, marital status was not found to be significantly associated with the outcome of the syphilis test among LDTDs (OR: 0.27, CI: 061– 5.9). 6.3.8.4 Syphilis Prevalence Based on Where LDTDs Sleep/Live during Road Trips Similar to the findings that the operation base of FSWs did not influence the chances of testing positive for syphilis, the residence of the truck drivers was not found to be significantly associated with testing positive for syphilis among This is expected, because 93% of all truckers lived/slept in their trucks. 45 6.3.8.5 Syphilis Prevalence Based on Age at Sex Debut and Exposure to Transactional Sex Similar to the findings on FSWs, the age at sex debut of an LDTD was not significantly associated with the outcome of syphilis test (OR: 0.28), CI: 0.034–2.23, Furthermore, the age at which an LDTD solicited for paid sex was also not related to the outcome of the syphilis test. 6.3.8.6 Syphilis Prevalence Based on Length of Service Similar to the findings on FSWs, length of service as an LDTD was also not associated with syphilis status in men (OR: 1.34, CI: 0.12–15.05). This suggests that new truckers are as equally exposed to syphilis as their older counterparts. 6.3.8.7 Syphilis Based on Exposure to Anal Sex Anal sex was not found to be a risk factors for syphilis acquisition among FSWs, LDTDs who had anal sex had apparently almost equal chances of acquiring syphilis infections as those who did not). It is therefore possible that FSWs, who are likely to play the receptive role in anal sex, are significantly more at risk that their male partners. 6.3.8.8 Syphilis Prevalence Based on Unprotected Sex and Condom Breaks An interesting observation was that FSWs who reported condom breaks were not found to be significantly more likely to test positive for syphilis than those who did not (OR: 2.56, CI: 0.87–7.54). Similar findings were made on LDTDs: Those who reported a condom break apparently had equal chances of testing positive for syphilis as those who did not (OR: 5.22, CI: 0.59–4.49). This is probably because condom use was relatively low among LDTDs. This presents a bigger risk that could mask the risk associated condom breakage. 6.3.8.9 Syphilis Prevalence Based on Alcohol The survey found that FSWs who take alcohol daily were not more likely to test positive for syphilis than those who had never taken alcohol (OR: 0.48, CI: 0.02–3.91). This is different from findings of alcohol use as a risk factor for syphilis acquisition among LDTDs. The reasons for these differences remain unclear, and further studies should be conducted to investigate this phenomenon. 6.3.8.10 Syphilis Prevalence Based on Knowledge of HIV/STIs and Self-Assessment of Possible Risk to Infection LDTDs who had previously been tested for HIV were equally likely to test positive as those who had not been tested before (OR: 1.26, CI: 0.14). These results are somehow expected considering that HIV testing was not voluntary among LDTDs and a majority of those tested before did it as a job requirement and did not collect their results. It is important therefore to formulate other related studies that can make voluntary HIV testing more acceptable among SWs and LDTDs. The chances of testing HIV positive were equal among LDTDs who believed that they were at risk of contracting HIV/syphilis and those who did not think so (OR: 0.51, CI: 0.05–5.02). Similar findings were made on FSWs and, therefore, self-assessment of risk is not an important factor that influences syphilis/HIV status. 6.3.9 Knowledge of HIV/AIDS and Its Prevention Strategies Respondents interviewed in this survey were asked whether they had heard of an illness called AIDS. Those who reported having heard of AIDS were asked a number of questions about whether and how AIDS could be prevented and how it is transmitted. Table 16 shows that awareness of HIV/AIDS 46 is high, with more than 93% of LDTDs having heard of AIDS. However, only 43% of the truckers were aware of the correct methods of preventing HIV infections. The prevention methods that were correctly identified by most LDTDs unassisted included safe blood transfusion (92%), having sex with only one uninfected female partner (88%), always using a condom while engaging in sexual intercourse (89%), avoiding sharing of injection equipment (86%) and abstaining from sexual intercourse (85%). However, various reported misconception of HIV prevention and transmission were reported among the truckers. About 31% of truckers who were aware of HIV/AIDS felt that one can prevent HIV infection by cleaning genitals after sex, by avoiding bites by mosquitoes or other insects (38%) and avoiding sharing of clothes or utensils (17%). Close to 72% of LDTDs believed that it is possible to guess if someone is HIV positive by physical assessment of by visual assessment Table 16. Percentage of LDTDs Reporting Knowledge of HIV/AIDS Ever heard of Ways a person can prevent becoming infected with HIV HIV/AIDS uninfected female partner Avoiding sharing injecting Having sex with only one mosquito or other insect Abstaining from sexual Not sharing clothes or Avoid being bitten by Ensuring safe blood Clean genitals after Using a condom intercourse intercourse transfusion equipment Know that utensils Background HIV/AIDS is characteristic No % preventable Age group < 20 25 88.0 45.5 100.0 100.0 70.0 40.0 0.0 80.0 80.0 40.0 20–24 136 96.3 42.7 89.1 89.1 82.0 34.0 14.0 94.1 94.0 28.0 25–29 116 94.0 50.5 77.8 82.7 86.0 41.9 21.0 91.2 95.4 44.2 30 34 89 92.1 41.5 88.2 91.2 85.7 44.4 11.1 85.7 85.7 25.0 35–39 75 89.3 41.8 78.6 92.9 87.0 39.1 17.4 78.3 91.3 21.7 40–44 30 93.3 50.0 92.9 92.3 100.0 25.0 16.6 75.0 92.0 25.0 45+ 55 92.7 29.4 80.0 86.7 76.9 23.1 23.1 92.3 93.3 30.8 Marital status Currently 236 91.1 41.4 83.1 87.4 82.4 39.7 20.5 86.9 63.5 38.4 married Divorced/ 57.1 separated 14 100.0 87.5 87.5 83.3 33.3 16.7 66.7 66.7 32.7 Widowed 42.9 Never 269 94.4 85.0 90.6 88.2 31.2 14.0 90.4 71.0 33.6 married 212 178 184 151 66 29 160 165 56 Total 526 93.2 (43.3) (84.8) (89.1) (85.5) (37.9) (16.9) (88.1) (92.2) (31.3) 6.3.9.1 Self-Assessment of Risk to HIV and Knowledge of HIV/AIDS Support Programs among LDTDs A high percentage (75%) reported having been tested for HIV previously. A majority (93%) of those tested indicated that the last test was voluntary, and 75% were counseled on HIV/AIDS and prevention methods (see Table 17). Slightly more than 30% of LDTDs were aware of drugs that can be used for the management of HIV/AIDS, and 16% had specific knowledge of ART. Half of those under the age of 20 were aware of ART. Knowledge of ART among other age groups was between 21% and 37%. More than 50% of the truckers knew of someone with AIDS or someone who had died 47 from AIDS. A small percentage (8%) indicated that they could tell if someone was HIV positive based by visual examination. More than 18% of LDTDs felt that they were at risk of contracting HIV, but 57% of all LDTDs did not consider themselves to be at risk. The rest did not know whether they were at risk or not. A small percentage (less than 10%) considered themselves at risk by engaging in anal sex, having more than five sex partners and having had a condom break during sex. Table 17. HIV Testing and Awareness of HIV/AIDS and Drugs That Help Treat People Who Have AIDS among Truckers by Socio-Demographic Characteristics HIV/AIDS Feel at risk of HIV management and Knowledge of HIV infection HIV testing (%) ART (%) infected persons (%) Knowledge of drugs for HIV/AIDS Know of someone with HIV/AIDS Can know someone’s HIV status prevention during last testing Counseled on HIV/AIDS and Last testing was voluntary by visual assessment Knowledge of ART or died from it Tested before management Background characteristics No. Percent Age group < 20 22 22.7 63.6 92.9 71.4 50.0 13.6 45.5 9.1 20–24 131 21.4 74.0 90.7 72.2 28.2 15.3 48.1 6.1 25–29 109 14.7 71.6 91.0 76.9 36.7 16.5 56.0 9.2 30–34 82 19.5 76.8 93.7 68.3 25.6 11.0 56.8 9.8 35–39 67 16.4 83.6 92.9 69.6 37.3 20.9 61.2 10.4 40–44 28 21.4 82.1 95.7 82.6 21.4 14.3 67.9 10.7 45+ 51 13.7 76.5 97.4 87.2 23.5 15.7 51.0 2.0 Marital status Currently married 215 14.4 79.1 93.5 79.4 28.4 14.4 55.6 7.0 Divorced/separated/ 14 28.6 78.6 81.8 72.7 28.6 21.4 64.3 14.3 widowed Never married 254 21.3 72.8 92.4 71.4 32.3 16.5 52.4 8.7 370 343 275 152 77 261 39 TOTAL 490 18.4 (75.8) (92.7) (75.1) (31.0) (15.7) (54.4) (8.1) 6.3.10 HIV Prevalence among LDTDs Overall HIV prevalence among truckers was estimated at 1%. The prevalence was highest among those aged over 44 years (3.7%). 48 Table 18. HIV Prevalence among LDTDs HIV status Percent HIV positive Number Age group < 20 0.0 24 20–24 0.8 133 25–29 0.9 107 30–34 0.0 86 35–39 1.4 69 40–44 0.0 30 45+ 3.7 54 Marital status Currently married 0.9 228 Divorced/separated/widowed 8.3 12 Never married 0.8 257 Total 1.0 503 6.3.10.1 HIV Prevalence Based on Age The HIV prevalence was highest for truckers aged 45 years and over and lowest for those aged between 30 and 34 and those aged between 40 and 44 years. Further analysis revealed that LDTDs over the age of 35 were more likely to test positive than those under the age of 35 years (OR: 3.48, CI: 0.47–30.0). The most probable explanation for this is that the LDTDs over the age of 35 are more likely to have engaged in risky behavior for longer than the younger truckers, but this explanation assumes that frequency and access to sex increases with age, which may not be the case for all LDTDs. Future studies should therefore address this observation. 6.3.10.2 HIV Prevalence Based on Literacy As mentioned before, the literacy levels among SWs and LDTDs were very low, and although a high proportion of LDTDs could read and write compared to that of the FSWs, a majority of truckers did not have more than 10 years of schooling. Interestingly, four (80%) of the five HIV-positive cases reported among LDTDs were found among those not able to read and write. Statistical tests did not, however, find any absolute relationship between literacy and HIV status among LDTDs (OR: 0.45, CI: 0.02–4.3). This is probably because, although the truckers may be better educated than their FSWs counterparts, they may not access critical information regarding HIV prevention because the majority are foreigners who are likely to encounter cultural and local language barriers. 6.3.10.3 HIV Prevalence Based on Marital Status We found a statistical differences between HIV prevalence among those never married and the divorced LDTDs (OR: 11.6, CI: 0.01–185) but no difference between those married and those unmarried (OR: 1.15, CI: 0.11–11.3). These data differ from the findings on FSWs, where marital status was a risk factor for infections. 49 6.3.10.4 HIV Prevalence Based on Accommodation of LDTDs All the HIV-positive cases among LDTDs were recorded among truckers who lived/slept in their trucks, but sleeping in trucks was not necessarily found to influence the HIV test of the entire population of truckers. 6.3.10.5 HIV Prevalence Based on Age at Sex Debut and Age at Exposure to Transactional Sex HIV prevalence among LDTDs who had a sex debut at the age of under 18 was not statistically different from that observed among those with a delayed debut (OR: 1.35, CI: 0.22–8.17). We also did not find any significant associations between the first ages a trucker sought transactional sex and HIV prevalence (OR: 0.96, CI: 0.17–1.17). Taken together, these results suggest that other factors could play a more critical role in HIV transmission and acquisition among the target groups than age at sex debut and age at first transactional sex. 6.3.10.6 HIV Prevalence Based on Type and Number of Sex Partners None of the HIV-positive cases were found among LDTDs with a defined number of partners and all positive cases were recorded among men who did not reveal how many partners they had. Similarly, all HIV-positive cases were found in LDTDs who did not reveal where they picked their sex partners. Because it was difficult to categorize such participants, it was not possible to run any statistical tests to analyze any associations between number of partners and HIV status 6.3.10.7 HIV Prevalence Based on Length of Service We were unable to show any relationships between the length of service of a trucker and the outcome of his HIV test. It is therefore possible that the risk of infections is not related to the length of time a trucker has worked. 6.3.10.8 HIV Prevalence Based on Exposure to Anal Sex Similar to previous findings that there was a correlation between anal sex and acquisition of HIV among FSWs (OR: 4.5, CI: 2.3–8.7), the LDTDs who reported to have engaged in anal sex were also found to be more likely to test positive for HIV than those who did not (OR: 9.2, CI: 1.49–57.03). These results further confirm that unprotected anal sex is a risk factor for HIV acquisition for LDTDs. 6.3.10.9 HIV Prevalence Based on Unprotected Sex and Condom Breaks The chances of testing positive for HIV were similar between LDTDs who reported to have used condoms and those who reported not using condoms in the last sexual engagement before recruitment to the survey. Similarly, no statistical differences were found between those who reported condom breaks and those who did not. The most probable explanation for these results is that the majority of LDTDs did not use condoms consistently and only about half of all truckers used condoms and, therefore, those who reported having used condoms in the last sexual encounter may have had unprotected sex in other sex acts. 6.3.10.10 HIV Prevalence Based on Use of Alcohol Statistical analysis failed to draw any significant associations between consumption of alcohol and HIV status among the LDTDs (OR: 0.24, CI: 0.014–3.909). These results show that alcohol use was not a significant risk for HIV acquisition in this population. 50 6.3.10.11 HIV Prevalence Based on Knowledge of HIV and Self-Assessment of Risk to HIV Infection All five HIV-positive cases were recorded among LDTDs who knew about the disease. However, statistical analysis revealed that HIV sero-positive status was not necessarily related to knowledge of HIV among LDTDs). As was the case among FSWs, self-assessment of risk to HIV was not found to be associated with the outcome of HIV test among the LDTDs (OR: 1.13, CI: 0.116–11.0). 51 7 DISCUSSION AND CONCLUSIONS This survey was designed to provide baseline data on behavioral risk factors for HIV and STIs. The survey also provides estimates on the prevalence of HIV and syphilis among FSWs and among LDTDs in Djibouti Ville. Past studies have found these two groups as vulnerable populations (Lacerda et al. 1997). Studies conducted in other countries show that transport corridors play an important role in shaping disease epidemics, including HIV and STIs such as syphilis (Gomez et al. 2013). Related studies also show that SWs (male and females) must be included in HIV/STI prevention programs because these populations are not only vulnerable, but can act as an important source of new infections and probably even new HIV strains (due to the associated sexual activities with multiple partners) (Stringer et al. 2006). The survey was formulated to fill existing knowledge gaps on HIV prevalence among FSWs and LDTDs operating in Djibouti Ville. It provides baseline data on potential risk factors for HIV and STI infections in the survey areas in Djibouti Ville. While the data provided in this report may not be sufficient to make general conclusions of HIV/STI status in Djibouti as a country, the survey provides important clues on the prevalence and pattern of spread of these infections in two important vulnerable groups. The survey provides basic statistical analysis of the results obtained. However, considering the small sample sizes of participants from some survey sites (e.g. PK12), and the small proportions of participants that tested positive for HIV and/or syphilis, the statistical findings of this survey should be interpreted with caution in order to avoid bias. Based on these considerations, we were cautious not to subject the data to multivariate analysis and only provided simple statistical analysis based on comparing selected dichotomous variables. We however anticipate that future studies, especially case-control studies, will provide robust data that that qualify for rigorous statistical analyses that will in turn provide a clear picture on interactions of multiple variables related to HIV/STIs epidemiology. Our survey estimates the prevalence of HIV among the FSWs was 13% and 1% among the LDTDs. The prevalence of syphilis was estimated at 5% among FSWs and at 1.6% among the LDTDs. These results show that HIV prevalence among FSWs in Djibouti is at least six times higher than the national average, while that observed among LDTDs was lower than reported nationally. The prevalence values of HIV and syphilis among LDTDs were relatively lower compared to those reported in related studies. These findings are interesting because LDTDs have traditionally been labeled as a “core transmitter group.” A survey carried out along the Kenya-Uganda transport corridor revealed an HIV prevalence of up to 35% among LDTDs compared to the control group (Mbugua et al. 1995). It is therefore clear that the prevalence recorded among the LDTDs in our current survey is relatively low. There is a likelihood that the prevalence of HIV and syphilis in our survey population differs from those reported in other studies due to differences in confounding factors and difference in composition of the target groups. It is also important to note that this was a cross-sectional survey and we cannot rule out that follow-up studies may provide data that significantly differs from that presented in this report. These assumptions are based on observations that the volume and 52 composition of LDTDs and FSWs may vary seasonally due to changes in economy, weather, and other factors. It is also important to note that majority of LDTDs who participated in this survey were from Ethiopia. The data presented here therefore underlines the need for cross-border measures to reduce HIV prevalence among highly mobile vulnerable groups. The most important factors identified as possible risk factors for infection with HIV and/or syphilis among the FSWs and LDTDs can be summarized as follows: being over the age of 35; engaging in (unprotected) sex; being divorced, separated or widowed; and daily consumption of alcohol among FSWs (although not among LDTDs) and longer experience in commercial sex work. Previous studies have shown a correlation between HIV sero-positivity and older age, longer duration working as a driver, fewer visits per month to spouses and contacts with commercial SWs per month (Richard et al. 2007). However, most of these variables were not identified as risk factors in the current survey. Other studies showing a correlation between truckers’ age/length of service have been published in South Africa (Ramjee et al. 2002) and Kenya (Mbugua et al. 1995). Taken together, our results suggest that risk factors for HIV and syphilis acquisition among LDTDs in Djibouti are likely to be different from those encountered in other countries, but the reasons behind these differences remain to be investigated. The survey found that most FSWs operating from brothels were more likely to have worked as SWs for more than 10 years than those operating in bars and streets. Based on these data, it is possible to assume that working from a fixed facility (such as a brothel) offers a SW a more stable client base and personal security, thereby allowing these women to work for longer periods of time. On the other hand, it is possible that FSWs operating in “stable” environments are more likely to establish regular customers with whom they may develop mutual trust that eventually leads to unprotected sex that may in turn raise the chances of infections. This survey also revealed that literacy is low among FSWs and LDTDs. There is a possibility that illiteracy contributes to poor knowledge as far as protecting oneself from HIV and STIs is concerned. It is also possible that some LDTDs are likely to experience difficulties in communication because the majority are foreigners from Ethiopia. A combination of illiteracy, language and other cultural barriers could partially explain why most participants were not conversant with programs meant to benefit vulnerable groups, such as those offered by ROADS and other NGOs operating in the survey area. Low literacy levels and few years in formal education could also be related to the early sex debut, as two-thirds of the SWs had a sex debut under the age of 18. There is therefore need to improve literacy levels in these vulnerable groups. In addition, visual campaigns using billboards, TV and audio commercials in languages and contexts that these groups can understand would be important. This survey further suggests that, depending on the category of the participants, between 15% and 40% of sex encounters are without protection. Studies conducted elsewhere show that the risk for HIV infection is lower among SWs who use condoms consistently (Ye et al. 2014). This survey also reveals that consistent use of condoms is low (at only about 50%) among the FSWs and the LDTDs. This likely explains the reason that “not using condoms” was identified as an important risk factor among FSWs and LDTDs. Similar to studies conducted elsewhere; availability of condoms and cost of condoms was not identified as important reasons for not using them (Bayer et al. 2013). Similarly, refusal by clients was the most commonly cited reason for non-use of condoms among the FSWs. In 53 related studies in Ghana, women cited client refusal (73%), receiving a higher payment (33%) and client brutality (43%) as reasons for not using condoms (Asamoah-Adu 2001). It has also been shown that male attitudes and behaviors, as well as discrimination against women, are critical factors that contribute to men’s demand for unprotected paid sex, while brothel owners may ask the SWs not to use condoms (Atteraya 2014). There are also occasional reports by SWs of the sub-standard quality of condoms (Silver et al. 1998). A small fraction of FSWs and LDTDs in this survey reported having experienced condom breaks during sex. Future studies should therefore investigate the quality of the condoms in circulation in Djibouti and other reasons behind reported condom breaks. Health caregivers and lobby groups should also teach the target groups about the proper use of condoms. Our survey revealed that a majority of FSWs engaged with both regular and occasional clients. Taken together, the data from this survey suggest that the nature of sexual relationships and the frequency of interaction are similar regardless of the category and operation base of the FSWs. It is important to note that 18 (38%) of the 47 HIV sero-positive FSW participants had regular sex partners. The importance of these results is even more significant based on the observation that 48 FSWs that had regular clients indicated that they did not use condoms regularly. It is also important to note that a number of clients revealed that they had maintained stable sexual relationships with their regular clients over a period of 10 years. Such stable relationships are likely to raise the chances of contracting and transmitting HIV and other STIs due to possible complacency that may develop between the FSWs and their clients over a period of time. Our survey reveals that a significant proportion of FSWs have regular clients with whom they have maintained a sexual relationship for more than 10 years. On the other hand, the LDTDs have multiple sex partners, while others were married. Such relations are likely to expose partner to infections. A similar study in Benin found that HIV prevalence was twice as high among boyfriends of SWs, suggesting that stable, long-term sexual relationships with FSWs are important risk factors to HIV/syphilis infections (Rwenge 2013). The findings of this survey of sexual liaisons between FSWs and regular and occasional clients was similar to those reported in related studies. A survey conducted in Gambia showed that condom use was not related to SW characteristics per se, but rather to client characteristics and setting (Peitzmeier et al. 2014). Other studies have also found condom use varies by type of sexual partner, and those with regular clients or partners are less likely to use condoms (Swe et al. 2013). It is also important to note that self-reported use of condoms tends to have bias and there is a possibility that condom use is actually lower than what is reported in this survey. To avoid such bias, we propose that future studies should use biological test methods to ascertain the use of condoms. One such method is based on detection of prostate-specific antigen on vaginal swabs. The presence of prostate-specific antigen indicates that recent unprotected sex took place. Such strategies have been used in Madagascar and Kenya (Mose et al. 2013). Our survey found that literacy levels among FSWs and LDTDs were very low. This could be related to low condom use in these two groups. This is because related studies have found that, although the setting of sex work appears to be a critical determinant of condom use, other characteristics, such as the education level of SWs and clients, are also important predictors of condom use. Our findings 54 point to the need for concerted efforts to update FSWs’ and LDTDs’ knowledge on correct and consistent use of condoms. This survey also investigated GBV. Sexual clients, especially regular clients, were identified as the majority of perpetrators of such acts. Unlike in related studies, police or security agents were not identified as major perpetrators of these crimes. Out of the 74 GBV cases, about 50% (37) were reported to the police of whom 28 were of a sexual nature. It is also important to address the issues of GBV among FSWs, and an even greater emphasis should be put into building and sharpening negotiation skills of FSWs for safer sex using condoms. Other studies show that the most successful strategy to increase safe and consistent condom use involves use of peer or other health educators to reach out to target groups, providing condoms and promoting and increasing HIV/STI testing and treatment (Schneider et al. 2014). This survey reveals that clients requested anal sex from the FSWs but anal sex was found be low in among the survey participants (4 reported cases). These findings are similar to those reported in a study in South Africa that showed that although requests for anal sex was as high as 50%, only 2% of FSWs met these requests (Ramjee et al. 2002). Our survey also revealed no significant association between anal sex and acquisition of HIV and or syphilis. These findings are inconsistent with other studies that have found that SWs who had anal sex were 3.5 times more likely to acquire HIV. Some studies conducted in South Africa have indicated that substituting oral sex and non-penetrative sexual practices for sexual acts that have a higher HIV risk, such as anal and vaginal sex, is a potential strategy for reducing the risk to SWs (Richter et al. 2013). However, promoting safer sex is likely to produce more meaningful results than campaigning for alterative non-protected oral or non- penetrative sex. The survey shows that a majority of the truckers are from Ethiopia and a majority of them sleep in their trucks during the road trips. About a fifth of these truckers engage in sexual activities with multiple partners whom they pick along the way or from brothels. Of importance to note is that some of the truckers, including those that are married, engage in unprotected sex with multiple SWs. A majority of the LDTDs indicated that they have ever solicited for paid sex in the course of their work. Related studies show that LDTDs are economically better off than other clients of FSW and are thus able to solicit for sex, including unprotected sex from women along the transport corridor (Rakwar et al. 1999). Another survey found that at the South Africa-Zimbabwe border, the favorite clients for commercial SWs are LDTDs who pay in foreign currency. These clients are preferred because they often pay more than the usual fee per encounter (Ramjee et al. 2002). Other studies show that LDTDs and FSWs may use middlemen to solicit for safer clients. Using middlemen also ensures that the clients can honor their pledge to pay for sex (Gysels et al. 2001). Although this survey did not investigate such practices, future studies should establish the role of purchasing power and the role played by middlemen in soliciting “safe” partners, because such practices could reduce the chances of using condoms among LDTDs and among FSWs. It is important to note that knowledge of STI symptoms was low among FSWs and even lower among LDTDs. It is therefore possible that STIs were under-reported in this survey and especially among LDTDs. Health-seeking behavior for suspected HIV infections is low among both categories of participants. In most categories, less than 30% had good knowledge on HIV and STIs. Most STIs are 55 not reported and most respondents reported not seeking professional medical assistance. This survey showed that a majority of LDTDs and SWs are not conversant with NGOs and civil-based organizations involved in sexual health and related campaigns in this area. This is not unexpected considering the limited services in the survey areas. ROADS is currently the only one operating in Djibouti Ville, with activities just starting in the other quartiers apart from PK12. Such facilities and programs, including those provided by ROADS, have the potential to educate and provide support for recognizing and treating HIV and other STIs. The government of Djibouti remains the major provider of support care and management of HIV/STI issues. There is clearly a need for more partners to support the government and devise innovative methods and strategies for providing care and support Mapping is necessary to allow health care planners and outreach programs to identify areas and regions where more or less resources should be allocated. The city planners and lobby groups can also use maps for devising strategies for providing security for the vulnerable groups. For the first time, this survey provides GIS maps for the survey areas indicating the SWs’ hotspots, locations of health care, resource centers, schools and other support facilities. Our GIS data indicate that most facilities, though limited in number, are situated in areas where they can be easily accessed. The maps reveal that most facilities and especially the health care clinics are not equitably distributed and that some areas, such as PK12, have very few of these facilities to match the volume of SWs and LDTDs who interact there. Our data compares well with that published by Morris and Ferguson (2005) that found that facilities that may support HIV and syphilis programs are few and poorly distributed across 47 hotspots along the Kenya-Uganda highway. It is our hope the government of Djibouti and its partners use these maps to identify hotspots and to more effectively allocate resources, such as condoms, ART programs and prevention outreach in the survey area. Since condom use is still low among SWs particularly, the government and NGOs can use these maps to develop strategies for condom distribution. It is important that caution is taken publishing such maps to protect the SWs from possible harassment. Our maps are available on request by researchers and planners. 56 8 RECOMMENDATIONS 1. This survey provides baseline data that have not been available on various issues related to HIV and syphilis infections and prevalence in Djibouti. The report provides critical data that can be used for formulation of other related studies in future. There is a need for further research on areas and sub-populations not covered by this survey. A follow-on survey to evaluate interventions that will take between 2 and 5 years to implement will be designed based on findings of this survey. Qualitative research will provide an in-depth understanding of emerging issues 2. Prevention is the mainstay of the HIV response, and SWs constitute a key population for HIV and STI prevention programs. In addition, access to HIV treatment and care for those with HIV is important, as this can potentially reduce the onward transmission of HIV. Services must be seen in the context of the Joint United Nations Programme on HIV/AIDS (UNAIDS) three pillars of an effective, evidence-informed response to HIV and sex work: a. Pillar 1: Ensuring universal access to comprehensive HIV prevention, treatment, care and support. b. Pillar 2: Building supportive environments strengthening partnerships and expanding choices. c. Pillar 3: Reducing vulnerability and addressing structural issues. 3. Condoms are the single most effective available technology for reducing sexual transmission of HIV. Therefore, an increase in correct, consistent condom use should be promoted as the main strategy for reducing unprotected sex. These should conform to quality standards and be made available free or at low cost to SWs and their clients. 4. Each partner has a role to play in maximizing successful negotiation of condom use, including enforcing their use. Comprehensive human rights-based condom programming and promotion among SWs and clients, as well as owners of sex work establishments and other intermediaries, should be a priority 5. Contacts between health workers and SWs or clients should be used to promote and provide condoms. Building the condom negotiation skills of SWs is also a central component of consistent condom use promotion and should include empowering SWs to initiate discussion of condom use with clients. 6. LDTDs have traditionally been considered a special vulnerable group and a transmitter group for HIV and STIs. However, this survey found that the prevalence of HIV and syphilis among this group were below the national averages of Djibouti and Ethiopia, the native country of most LDTDs. Future studies should investigate the reasons behind these apparent discrepancies 7. Sex debut was early in the two groups. Early sex debut is a known risk factor for STI infections including those that may lead to development of cervical cancer. Future programs should address these issues and probably advocate for sex education in schools. In order for these groups to benefit, sex education should start early because majority of participants had a less than 10-year schooling. 57 8. Majority of LDTDs who participated in this survey were from Ethiopia. There is therefore a need to launch cross-border HIV/STI prevention and sensitization programs targeting such highly mobile and vulnerable groups. Future studies should investigate whether LDTDs, the majority of whom are foreigners, face language and cultural barriers that hinder their access to critical information on HIV and syphilis prevention. 9. It is also important to consider providing HIV/STI information in form of bill boards and pamphlets in language that the truckers would include alongside the national language on Djibouti. 10. This survey suggests that various factors such as anal sex, long period of engagement as a sex worker, alcohol consumption among others may be significant risk factor for HIV/STI acquisition in one or both groups studied. While the data is not conclusive due to the small sample size of positive participants and the fact that only selected sites were included, future studies, especially case-control studies should be designed in future with a view to providing a deeper understanding on interaction of multiple confounding factors. 11. A significant proportion of participants did not consider themselves to be at risk of HIV/STIs despite engaging in risky behavior such as unprotected anal and genital sex. Future programs should enlighten target groups on how to accurately assess their individual vulnerabilities. Such campaigns should include lobbying for frequent HIV/STI testing. 12. Condoms are the single most effective available technology for reducing sexual transmission of HIV. Therefore, an increase in correct, consistent condom use should be promoted as the main strategy for reducing unprotected sex. These should conform to quality standards and be made available free or at low cost to SWs and their clients. The use of condoms is significantly low in both groups. There is a need to promote safe sex especially between the FSWs and the regular clients with whom they often develop long-term relationships. Sensitization campaigns on safe sex should also be supplemented with accurate information on family planning methods. Each partner has a role to play in maximizing successful negotiation of condom use, including enforcing their use. Comprehensive human rights- based condom programming and promotion among SWs and clients, as well as owners of sex work establishments and other intermediaries, should be a priority 13. A curious finding in this survey was the lack of an obvious link between use of injectable drugs, sharing needles or condom breakage and HIV/Syphilis status. Future case-control studies should verify these findings. 14. It is important to note that both groups largely comprised of young participants (less than 25 years old). Considering that this survey only recruited adults (above the age of 18 years), there are possibilities that children (under the age of 18 years) may also be working as sex workers and truckers. Future studies and programs should also focus on these groups and find strategies to reach to younger ages through media including social media 15. There is a need to encourage FSWs and LDTDs to know their HIV/syphilis status and measures must be put in place to improve access to HIV care and support (including access to ART) among those who turn positive. The government of Djibouti and her partners can improve the efficient of service provision and support for vulnerable groups by using GIS maps such as those generated in this survey. 58 9 APPENDICES Appendix 1. Statistical Tests for HIV and Syphilis prevalence by Various Characteristics among FSWs HIV HIV Syphilis Syphilis Test category/variable positive negative CI OR positive negative CI OR Could read or write 12 62 2 81 0.64–2.98 1.39 0.06-1.09 0.40 Could not read and write 34 244 16 262 Over Age > 30 years 33 83 7 47 3.20–13.62 6.56 1.49–10.9 4.04 Age of ≤ 30 years 14 231 11 298 Over age of 35 years 17 37 7 47 0.47–2.97 1.18 0.55–3.65 1.41 Between 25 and 34 years 23 144 16 151 divorced/separated 26 99 10 230 1.18–4.77 2.36 0.23–01.56 0.60 Not married 18 162 8 110 Married 1 4 10 230 0.4–3.4 0.36 0.29–9.68 0.50 Not Married 20 220 0 5 >5 clients per week 17 140 6 151 0.31–1.4 0.65 0.16–1.53 0.49 ≤ 5 clients per week 15 80 7 88 On contraceptives 34 222 14 221 0.57–2.45 1.15 0.53–9.28 2.07 Not on contraceptive 12 90 3 98 Sex debut >18 years 23 142 11 185 0.60–2.22 1.15 0.51–3.59 1.35 Sex debut ≤ 18 years 24 1.71 7 160 Debut for commercial sex > 18 years 42 248 14 276 0.81–6.7 2.24 0.25-3.17 0.85 Debut for commercial sex ≤ 18 years 5 66 4 67 Operate from public places (e.g., 23 205 8 220 bars) 0.33–1.13 0.61 0.17–1.15 0.44 Operate from brothels or homes 24 131 10 121 Takes alcohol daily 8 88 4 92 0.16–0.94 0.39 0.16-2.12 0.61 Never takes alcohol 34 147 12 169 Uses drugs 14 104 3 117 0.41–1.71 0.84 0.09–1.47 0.39 Does not use drugs 33 210 15 223 With regular clients 41 268 16 293 0.73–1.9 1.18 0.47–1.95 0.96 With occasional clients 34 263 16 281 Drug user, shared needles 2 27 2 27 0.08–2.15 0.40 0.04–1.29 0.26 Drug user, did not share needles 11 60 16 55 Experienced GBV 7 67 1 73 0.25–1.60 0.65 0.01–1.84 0.25 Never experienced GBV 40 249 15 273 59 HIV HIV Syphilis Syphilis Test category/variable positive negative CI OR positive negative CI OR Knows correct prevention of HIV 31 217 12 236 Doesn’t know correct prevention of 0.19–0.75 0.37 0.16–1.39 0.44 16 42 6 52 HIV Experienced condom break during 9 36 5 45 sex 0.66–3.77 1.59 0.77–8.62 2.65 Never experienced condom break 38 262 12 286 Did not use condoms (last 9 43 2 50 intercourse) 1.20–7.68 3.066 0.11–3.31 0.69 Used condoms (last intercourse) 20 293 16 279 Operating from Arhiba and 42 253 15 280 Quartier 2 0.73-6.09 3.03 0.29-5.05 1.13 Operating from elsewhere 5 61 3 63 Worked for > 3 years as a SW 29 132 12 184 1.15-4.42 2.25 0.72-6.52 2.13 Worked for ≤ 3 years as a SW 18 184 6 196 Feels at risk of HIV 26 118 5 139 0.99–5.05 2.22 0.14–1.7 0.46 Doesn’t know if at risk of HIV 11 111 7 90 The table shows bivariate tests for dichotomous variables. Chi-tests were used except in significantly small values where Fisher exact tests were used. Participants who answered “don’t know” and those who declined to answer a question on a particular variable were excluded from the test. Confidence intervals were set at 95% cutoff. 60 10 REFERENCES AND FURTHER READING 1. Asamoah-Adu C, Khonde N, Avorkliah M, Bekoe V, Alary M, Mondor M, Frost E, Deceuninck G, Asamoah-Adu A, Pépin J (2001). “HIV infection among sex workers in Accra: need to target new recruits entering the trade.” J Acquir Immune Defic Syndr. 28(4):358–66. 2. Atteraya MS, Kimm H, Song IH (2014). “Women’s autonomy in negotiating safer sex to prevent HIV: findings from the 2011 Nepal Demographic and Health Survey.” AIDS Educ Prev 26(1):1–12. 3. Barongo LR, Borgdorff MW, Mosha FF, et al. (1992). “The epidemiology of HIV-1 infection in urban areas, roadside settlements and rural villages in Mwanza Region, Tanzania.” AIDS 6 (12): 1521–8. 4. Bayer AM, Garvich M, Díaz DA, Sánchez H, García PJ, Coates TJ (2013). “When Sex Work Becomes Your Everything: The Complex Linkages Between Economy and Affection Among Male Sex Workers in Peru.” Am J Mens Health. [Epub ahead of print]. 5. Bergsjo P, Olomi RM, Talle A, Klepp KI (1995). “Bar attendants as health educators. Prevention of sexually transmitted diseases in high-risk areas.” Tidsskr Nor Laegeforen 115(26): 3281–3. 6. Bikaako W (2000). AIDS and transport: the experience of Ugandan road and rail transport workers and their unions. London: International Transport Workers’ Federation (ITF). 7. Brushett S, Osika J (2005). Lessons learned to date from HIV/AIDS transport corridor project. World Bank Global HIV/AIDS Program Discussion Paper. Washington: World Bank. 8. Bwayo JJ, Mutere AN, Omari MA, et al. (1991). “Long-distance truck drivers. 2: Knowledge and attitudes concerning sexually transmitted diseases and sexual behaviour.” East African Medical Journal 68 (9): 714–9. 9. Bwayo JJ, Plummer F, Omari MA, et al. (1994). “Human immunodeficiency virus infection in long-distance truck drivers in east Africa.” Archives of Internal Medicine 154 (12): 1391–6. 10. Carswell JW, Lloyd G, Howells J (1989). “Prevalence of HIV-1 in east African lorry drivers.” AIDS 3(11): 759–61. 11. Family Health International (2003). Behavioural surveillance survey. Zambia Long distance truck drivers, light truck and minibus drivers and uniformed personnel in transportation border routes. Arlington, VA: Family Health International. 12. Family Health International (2001). Lesotho and Swaziland: HIV/AIDS risk assessments at cross borders and migrants sites in Southern Africa. Arlington, VA: Family Health International. 13. Gomez GB, Venter WDF, Lange JMA, Rees H, and Hankins C, “North-South Corridor Demonstration Project: Ethical and Logistical Challenges in the Design of a Demonstration Study of Early Antiretroviral Treatment for Long Distance Truck Drivers along a Transport Corridor through South Africa, Zimbabwe, and Zambia,” Advances in Preventive Medicine, vol. 2013. doi:10.1155/2013/190190 61 14. Gysels M, Pool R, Bwanika K (2001). “Truck drivers, middlemen and commercial sex workers: AIDS and the mediation of sex in southwest Uganda.” AIDS Care 13(3): 373–85. 15. Gysels M, Robert P, and Bwanika K (2001). “Truck drivers, middlemen and commercial sex workers: AIDS and the mediation of sex in south west Uganda.” AIDS Care 13 (3) 373–385. 16. Hoffmann O, Zaba B, Wolff B, et al. (2004). “Methodological lessons from a cohort study of high-risk women in Tanzania.” Sexually Transmitted Infections 80: 69–73. 17. International Organization for Migration (IOM) and Joint United Nations Programme on HIV/AIDS (UNAIDS) (2005). HIV and Mobile workers: a review of risk and programmes among truckers in West Africa. Geneva: IOM/UNAIDS. 18. IOM (2003). Mobile populations and HIV/AIDS in the Southern Africa Region. Recommendations for Action. Desk review and Bibliography on HIV/AIDS and Mobile Populations. Pretoria: IOM. 19. IOM/CARE (2003). Mobility and HIV/AIDS in Southern Africa: a field study in South Africa, Zimbabwe and Mozambique. Pretoria: IOM/CARE. 20. Jackson DJ, Rakwar JP, Richardson BA (1997). “Decreased incidence of sexually transmitted diseases among trucking company workers in Kenya: results of a behavioral risk-reduction programme.” AIDS 11(7): 903–9. 21. Job B, Omari MA, Mutere AN, Plummer FA, et al. (1992). “HIV infection in long distance truck drivers in Kenya: seroprevalence, seroincidence, and risk factors.” International Conference on AIDS Jul 19–24 8: Th68 (abstract no. ThC1514). 22. Kilimwiko L (1991). “Condoms hitch lift with truckers.” WorldAIDS (17): 9. 23. Lacerda R, Gravato N, McFarland W, Rutherford G, Iskrant K, Stall R, Hearst N (1997). “Truck drivers in Brazil: prevalence of HIV and other sexually transmitted diseases, risk behavior and potential for spread of infection.” AIDS Sep;11 Suppl 1:S15–9. 24. Laukamm-Josten U, Mwizarubi BK, Ouwater A, et al. (2000). “Preventing HIV infection through peer education and condom promotion among truck drivers and their sexual partners in Tanzania, 1990–1993.” AIDS Care 12(1): 27–40. 25. Lee S (2004). The vulnerability of women street traders to HIV/AIDS: local government opportunities for action. Durban: University of KwaZulu-Natal/HEARD. 26. Marcelin, Anne‐Geneviève, et al. (2002) "Kaposi's sarcoma herpesvirus and HIV‐1 seroprevalences in prostitutes in Djibouti." Journal of medical virology 68(2):164-167. 27. Marck J (1999). “Long-distance truck drivers’ sexual cultures and attempts to reduce HIV risk behavior amongst them: a review of African and Asian literature.” In Resistances to behavioural changes to reduce HIV AIDS infection in predominantly heterosexual epidemics in Third World Countries. Edited by Caldwell, JC, Caldwell P, Anarfi K, et al. Canberra: Health Transition Center. 28. Mbugua GG, Muthami LN, Mutura CW, et al. (1995). “Epidemiology of HIV infection among long distance truck drivers in Kenya.” East African Medical Journal 72 (8): 515–8. 62 29. Mbugua GG, Muthami LN, Mutura CW, Oogo SA, Waiyaki PG, Lindan CP, Hearst N (1995). “Epidemiology of HIV infection among long distance truck drivers in Kenya.” East African Medical Journal [1995, 72(8):515–8. 30. Morris C, Ferguson A (2005). Hot spot mapping of the Northern corridor transport route: Mombasa-Kampala. Final report. Nairobi: University of Nairobi/University of Manitoba. 31. Morris, CN and Alan GF (2006). “Estimation of the sexual transmission of HIV in Kenya and Uganda on the trans-Africa highway: the continuing role for prevention in high risk groups.” Sexually transmitted infections 82(5) 368–371. 32. Mose F, Newman LP, Njunguna R, Tamooh H, John-Stewart G, Farquhar C, Kiarie J (2013). “Biomarker evaluation of self-reported condom use among women in HIV-discordant couples.” Int J STD AIDS 24(7):537–40. 33. Muizarubi B, Cole L, Outwater A, Lamson N (1991). “Targeting truckers in Tanzania. Prevention and awareness.” AIDS and Society 4–5. 34. Mupemba K (1999). “The Zimbabwe HIV prevention program for truck drivers and commercial sex workers: a behaviour change intervention.” In Resistances to behavioral changes to reduce HIV/AIDS infection in Predominantly heterosexual epidemics in Third World Countries. Edited by Caldwell, JC, Caldwell P, Anarfi K, et al. Canberra: Health Transition Center. 35. Ntozi JP, Najjumba IM, Ahimbisiwe F, et al. (2003). “Has the HIV/AIDS epidemic changed sexual behaviour of high-risk groups in Uganda?” African Health Sciences 3 (3): 107–16. 36. Nunn AJ, Wagner HU, Kamali A, Kengeya-Kayondo JF, et al. (1995). “Migration and HIV-1 seroprevalence in a rural Ugandan population.” AIDS 9(5): 503–6. 37. Nunn AJ, Wagner HU, Okongo JM, et al. (1996). “HIV-1 infection in a Ugandan town on the trans African highway: prevalence and risk factors.” International Journal of STD and AIDS 7(2): 123–30. 38. Nyamuryekung’e K, Laukamm-Josten U, Vuylsteke B, et al. (1997). “STD services for women at truck stop in Tanzania: evaluation of acceptable approaches.” East African Medical Journal; 74 (6): 343–7. 39. Nzyuko S, Lurie P, McFarland W, et al. (1997). “Adolescent sexual behaviour along the Trans Africa Highway in Kenya.” AIDS 11 Suppl. 1: S21–6. 40. Peitzmeier S, Mason K, Ceesay N, Diouf D, Drame F, Loum J, Baral S (2014). “A cross-sectional evaluation of the prevalence and associations of HIV among female sex workers in the Gambia.” Int J STD AIDS 2;25(4):244–52. 41. Pickering H, Okongo M, Nnalusiba B, et al. (1997). “Sexual networks in Uganda: casual and commercial sex in a trading town.” AIDS Care 9(2): 199–207. 42. Rakwar J, Lavreys L, Thompson ML, Jackson D, Bwayo J, Hassanali S, Mandaliya K, Ndinya- Achola J, Kreiss, J (1999). “Cofactors for the acquisition of HIV-1 among heterosexual men: prospective cohort study of trucking company workers in Kenya.” AIDS 13(5): 607–614. 63 43. Ramjee G, Gouws E (2002). “MPH Prevalence of HIV Among Truck Drivers Visiting Sex Workers in KwaZulu-Natal, South Africa.” Sexually Transmitted Diseases 29: (1)44–49. 44. Richard L, Victor P, Florence W (2007). “Migration, HIV/AIDS Knowledge, Perception of Risk and Condom Use in the Senegal River Valley.” International Studies in Population Volume 6, 2007, pp. 171–194. 45. Richter ML, Chersich M, Temmerman M, Luchters S (2013). “Characteristics, sexual behavior and risk factors of female, male and transgender sex workers in South Africa.” S Afr Med J. 19;103(4):246–51. 46. Riedner G, Hoffmann O, Rusizoka M, et al. (2006). “Decline in sexually transmitted infection prevalence and HIV incidence in female barworkers attending prevention and care services in Mbeya Region, Tanzania.” AIDS 20(4): 609–615. 47. Ritter T (2000). Commercial sex workers and truck drivers: background for a core transmitter intervention. Uganda: Delivery of Improved Services for Health (DISH II) 48. Robinson, E. T. "Reaching men: at work and in social settings." Network 12.1 (1991): 9-15. 49. Rwenge JR (2013). “Sexual behavioral among adolescents and young people in sub-Saharan Africa and related factors.” Afr J Reprod Health 17(1):49–66. 50. Schneider JA, Zhou AN, Laumann EO (2014). “A new HIV prevention network approach: Sociometric peer change agent selection.” Soc Sci Med. S0277-9536(14)00044–6. 51. Silver LD, Castilho SR (1998). “Evaluating technologies in reproductive health: case studies of a consumer protection approach.” Int J Gynaecol Obstet. 63 Suppl 1:S75–83. 52. Stringer JS, Zulu I, Levy J, Stringer EM, Mwango A, Chi BH, Mtonga V, Reid S, Cantrell RA, Bulterys M, Saag MS, Marlink RG, Mwinga A, Ellerbrock TV, Sinkala M (2006). “Rapid scale-up of antiretroviral therapy at primary care sites in Zambia: feasibility and early outcomes.” JAMA 16;296(7):782–93. 53. Swe LA, Rashid A. (2013). “HIV prevalence among the female sex workers in major cities in Myanmar and the risk behaviors associated with it.” HIV AIDS (Auckl). 28;5:223–30. 54. The 2013 CIA World Fact Book 55. The Synergy Project/University of Washington (2002). Putting on the brakes. Preventing HIV transmission along truck routes. Washington: University of Washington. 56. UNDP (2011) Mobility, Migration and HIV Vulnerability of Populations along the Ports of the Red Sea and Gulf of Aden - Situation and Response Analysis report. Www.undp.org/content/dam/undp/library/hivaids/Mobility%20Migration%20and%20HIV%20V ulnerability%20Report%202012.pdf 57. UNAIDS (2001). Technical update: Population Mobility and AIDS. UNAIDS Best Practice Collection. Geneva: UNAIDS.UNAIDS - Djibouti 2012 Country Progress Report 58. UN Data (2013): World Statistics Pocketbook 64 59. Université de Johns Hopkins Centre pour les Programmes de Communication (JHU/CCP) (2000). Définir les voies pour la prévention du VIH/SIDA au Rwanda: Leçons apprises sur les aspects comportementaux. Baltimore, MD: JHU/CCP. 60. Ye L, Wei S, Zou Y, Yang X, Abdullah AS, Zhong X, Ruan Y, Lin X, Li M, Wu D, Jiang J, Xie P, Huang J, Liang B, Zhou B, Su J, Liang H, Huang A (2014). “HIV pre-exposure prophylaxis interest among female sex workers in Guangxi, China.” PLoS One 22;9(1):e86200. 61. Zuma K, Gouws E, Williams B, Lurie M (2003). “Risk factors for HIV infection among women in Carltonville, South Africa: migration, demography, and sexually transmitted diseases.” International Journal for STD and AIDS 14: 814–17. 65