University of Ghana http://ugspace.ug.edu.gh SCHOOL OF PUBLIC HEALTH COLLEGE OF HEALTH SCIENCES UNIVERSITY OF GHANA VIOLENCE AGAINST FEMALE SEX WORKERS IN GHANA: ANALYSIS OF ASSOCIATED FACTORS BY PAUL AYAMAH (10701132) A DISSERTATION SUBMITTED TO THE SCHOOL OF PUBLIC HEALTH, UNIVERSITY OF GHANA IN PARTIAL FULFILLMENT FOR THE AWARD OF MASTER OF PUBLIC HEALTH (MPH) DEGREE JULY, 2019 University of Ghana http://ugspace.ug.edu.gh DECLARATION I, Paul Ayamah, declare that, except for other people’s investigations which have been duly acknowledged, this dissertation is the product of my original research work written under the supervision of Dr. Justice Moses K. Aheto of the Department of Biostatistics, School of Public Health, and that this dissertation, either in whole or in part has not been presented elsewhere for another degree. 30th September, 2019 …………………………….. ………………………………………. Paul Ayamah Date (Student) 30th September, 2019 ……………………………….. ……………………………………….. Dr. Justice Moses K. Aheto Date (Supervisor) i University of Ghana http://ugspace.ug.edu.gh DEDICATION To my mother, Ms. Sarah Mbayaab ii University of Ghana http://ugspace.ug.edu.gh ACKNOWLEDGEMENT My greatest gratitude to Christ, the fount of all I have and the hope of my aspirations. Granted a million opportunities, I would always want to be under his supervision! Dr. Justice Moses K. Aheto, I am sincerely grateful for the priceless effort you invested in shaping this work. I also wish to thank the management of Ghana AIDS Commission for the permission to use the 2015 IBBSS data for this study. I am equally full of gratitude to Mr. Edward Adiiboka and Ms. Rita Afriyie for their assistance at the initial stages of this journey, Mr. David Nartey Tetteh a fellow soldier and great companion. Finally, to family and friends who supported me in diverse ways in this academic journey – may the good Lord replenish your investments bountifully. iii University of Ghana http://ugspace.ug.edu.gh ABSTRACT Background: Violence against female sex workers (FSWs) is a widespread phenomenon induced primarily by the pervasive worldview that sex work is a contravention of morality. It is however necessary to dismantle the barrier posed by violence against FSWs in order to reach them with HIV services due to the pivotal role they, their clients and partners of their clients play in reinforcing the HIV epidemic. Objective: This study sought to assess the risk factors for violence against FSWs in Ghana. Method: The study used secondary data from the 2015 IBBSS among FSWs (n=4,279) and their non-paying partners conducted by Ghana AIDS Commission (GAC). Bivariate, and multiple logistic regression analyses were done to establish factors associated with violence against FSWs. Statistical significance was set at a p- value<0.05. Results: The study established the prevalence of physical and sexual violence as 10.12% and 24.75% respectively. Physical violence was perpetrated mainly by one- time paying clients (50.58%) and regular paying clients (30.95%). A similar trend was reported for sexual violence. As established by the study, the predictors of physical violence are type of FSW (p-value<0.01), current age of FSW (p-value=0.02), region (p-value<0.01), whether FSW has ever used drugs (<0.01) and whether FSW had ever used alcohol before sex in the past six months preceding the survey (p-value<0.01). Sexual violence on the other hand was predicted by current age of FSW (p-value=0.02), marital status (p-value<0.01), number of living children (p-value=0.04), where FSWs take clients for sex (p-value <0.01), region (p-value<0.01), whether FSW has a history of drug use (p-value <0.01), and whether FSW has ever used alcohol before sex in the past six months (p-value<0.01). Both physical and sexual violence had negative consequences on consistency of condom use and condom failure among FSWs. Conclusion: Violence against FSWs is pervasive in Ghana and it is influenced by a myriad of factors. There is the need for tailored interventions to address the menace in order to reach FSWs effectively with HIV services. iv University of Ghana http://ugspace.ug.edu.gh TABLE OF CONTENTS DECLARATION ...................................................................................................................... i DEDICATION ......................................................................................................................... ii ACKNOWLEDGEMENT ...................................................................................................... iii ABSTRACT ............................................................................................................................. iv TABLE OF CONTENTS ........................................................................................................ v LIST OF TABLES ................................................................................................................ viii LIST OF FIGURES .............................................................................................................. viii LIST OF ABBREVIATIONS ................................................................................................ ix CHAPTER ONE ...................................................................................................................... 1 INTRODUCTION.................................................................................................................... 1 1.1 Study background ................................................................................................................ 1 1.2 Statement of the problem ..................................................................................................... 4 1.3 Research questions ............................................................................................................... 6 1.4 Study objectives ................................................................................................................... 6 1.4.1 General objective .......................................................................................................... 6 1.4.2 Specific Objectives ....................................................................................................... 6 1.5 Justification .......................................................................................................................... 6 1.6 Theoretical Framework: The Social Ecological Model ....................................................... 8 1.7 Conceptual framework: Correlates of FSWs’ vulnerability to violence .............................. 9 1.7.1 Narrative on conceptual framework ............................................................................ 10 CHAPTER TWO ................................................................................................................... 11 LITERATURE REVIEW ..................................................................................................... 11 2.0 Introduction ........................................................................................................................ 11 2.1 Summary of the 2015 IBBSS ............................................................................................. 11 2.2 Overview of sex work in Ghana ........................................................................................ 12 2.3 Gender-based violence and risk of HIV acquisition and/or transmission .......................... 14 2.4 Sources and Types of Violence against Female Sex Workers ........................................... 15 2.5 Determinants of FSWs susceptibility to violence .............................................................. 16 2.5.1 Occupational factors that heighten FSWs’ vulnerability to violence .......................... 16 2.5.1.1 Condom negotiation ............................................................................................. 16 2.5.1.2 Servicing clients in open spaces ........................................................................... 17 2.5.1.3 Servicing higher number of clients ...................................................................... 18 2.5.1.4 Duration in sex work ............................................................................................ 18 2.5.2 Personal ....................................................................................................................... 19 2.5.2.1 Age ....................................................................................................................... 19 2.5.2.2 Knowledge and perception of violence among FSWs ......................................... 20 v University of Ghana http://ugspace.ug.edu.gh 2.5.2.3 Drug and alcohol use ........................................................................................... 20 2.5.2.4 Being HIV positive .............................................................................................. 21 2.5.2.5 Financial insecurity .............................................................................................. 21 2.5.2.6 Residency (migrant or local dweller) ................................................................... 21 2.5.2.7 Educational status ................................................................................................ 22 2.5.3 Legal/Structural........................................................................................................... 22 2.5.3.1 Law enforcement ................................................................................................. 22 2.5.4 Other factors................................................................................................................ 23 2.5.4.1 Gender inequality ................................................................................................. 23 2.5.4.2 Homelessness ....................................................................................................... 24 2.5.4.3 Societal anti sex work sentiments ........................................................................ 25 2.5.4.4 Sex trafficking, contract sex work and related forms of sex work ....................... 25 2.5.4.5 Miscellaneous factors ........................................................................................... 26 2.6 Factors influencing HIV acquisition and transmission among FSWs ............................... 27 2.6.1 Multiple sexual partners .......................................................................................... 27 2.6.2 High risk behaviour of clients ................................................................................. 27 2.6.3 Drug and alcohol use .............................................................................................. 28 2.6.4 Experience of violence ............................................................................................ 28 2.6.5 Presence of other STI .............................................................................................. 29 2.6.6 Low condom use with non-paying partners and regular clients.............................. 29 2.6.7 Early sex debut and duration in sex work ............................................................... 30 2.7 Consequence of violence against female sex workers ....................................................... 30 2.7.1 Risk of HIV and other STI infections ..................................................................... 30 2.7.2 Increased migration ................................................................................................. 31 2.7.3 Negative mental health outcomes ........................................................................... 31 2.7.4 Adverse Reproductive Health Outcome ................................................................. 32 2.7.5 Miscellaneous consequences ................................................................................... 32 2.8 HIV as a public health threat ............................................................................................. 33 2.9 Conclusion ......................................................................................................................... 34 CHAPTER THREE ............................................................................................................... 35 METHOD ............................................................................................................................... 35 3.0 Introduction .................................................................................................................... 35 3.1 Study Design .................................................................................................................. 35 3.2 Study Location ............................................................................................................... 36 3.2.1 Re-demarcation of Regions in Ghana ..................................................................... 36 3.3 Variables ........................................................................................................................ 37 vi University of Ghana http://ugspace.ug.edu.gh 3.3.1 Dependent Variables ............................................................................................... 37 3.3.2 Independent Variables ............................................................................................ 38 3.4 Study Population ............................................................................................................ 40 3.5 Preparation of data for analysis ...................................................................................... 40 3.6 Bivariate Analysis .......................................................................................................... 41 3.7 Simple and multiple logistic regression analysis ........................................................... 41 3.8 Ethical Consideration ..................................................................................................... 42 CHAPTER FOUR .................................................................................................................. 43 RESULTS ............................................................................................................................... 43 4.0 Introduction ........................................................................................................................ 43 4.1 Demographic Characteristics of FSWs .............................................................................. 43 4.2 Sources of violence against FSWs ..................................................................................... 45 4.3 Bivariate analysis of factors associated with physical violence against FSWs .................. 46 4.4 Bivariate analysis of factors associated with sexual violence against FSWs ..................... 48 4.5 Predictors of physical violence against FSWs ................................................................... 51 4.6 Predictors of sexual violence against FSWs ...................................................................... 53 4.7 Consequences of violence against FSW ............................................................................ 55 CHAPTER FIVE ................................................................................................................... 58 DISCUSSION ......................................................................................................................... 58 5.0 Introduction ........................................................................................................................ 58 5.1 Socio-demographic characteristics .................................................................................... 58 5.2 Sources of violence against FSWs ..................................................................................... 59 5.3 Predictors of violence against FSWs ................................................................................. 60 5.4 Consequences of violence against FSWs ........................................................................... 61 CHAPTER SIX ...................................................................................................................... 63 CONCLUSIONS AND RECOMMEDATIONS .................................................................. 63 6.0 Introduction ........................................................................................................................ 63 6.1 Conclusions ........................................................................................................................ 63 6.2 Recommendations .............................................................................................................. 63 6.3 Limitations ......................................................................................................................... 65 REFERENCES ....................................................................................................................... 66 vii University of Ghana http://ugspace.ug.edu.gh LIST OF TABLES Table 1. Demographic characteristics ...................................................................................... 44 Table 2: Bivariate analysis of factors associated with physical violence against FSWs ......... 47 Table 3. Bivariate analysis of factors associated with sexual violence against FSWs ............. 49 Table 4. Multiple logistic regression analysis of factors influencing physical violence against FSWs ........................................................................................................................................ 51 Table 5. Multiple logistic regression analysis of factors influencing sexual violence against FSWs ........................................................................................................................................ 54 Table 6. Analysis of violence against FSWS and consistent condom use with paying clients 56 Table 7. Analysis of violence against FSWs and condom failure ............................................ 56 Table 8. Analysis of violence against FSWS and HIV testing behaviour................................ 57 Table 9. Analysis of violence against FSWS and HIV test outcome ....................................... 57 LIST OF FIGURES Figure 1. Conceptual Framework: Risk factors for violence against FSWs .............................. 9 Figure 2. Map of Ghana (before 2018 Referendum) .............................................................. 37 Figure 3. Map of Ghana (after the 2018 referendum) .............................................................. 37 Figure 4. Prevalence, and sources of violence against FSWs .................................................. 46 viii University of Ghana http://ugspace.ug.edu.gh LIST OF ABBREVIATIONS AIDS Acquired Immunodeficiency Syndrome FHI Family Health International FSW Female Sex Worker GAC Ghana AIDS Commission GBV Gender-based Violence HIV Human Immunodeficiency Virus IBBSS Integrated Bio-Behavioural Sentinel Survey ICVS International Crime Victims Survey NMIMR Noguchi Memorial Institute for Medical Research Non-PP Non-Paying Partner NSP National HIV and AIDS Strategic Plan PHSC Protection of Human Subject Committee PLHIV Person(s) Living with HIV SDG Sustainable Development Goals STI Sexually Transmitted Infections UNAIDS The Joint United Nations Programme on HIV/AIDS VAW Violence against Women WHO World Health Organisation ix University of Ghana http://ugspace.ug.edu.gh CHAPTER ONE INTRODUCTION 1.1 Study background Violence against female sex workers (FSWs) is a widespread phenomenon (Berger et al., 2018; Decker et al., 2015; Fawole, & Dagunduro, 2014; Platt et al., 2011) because the notion that sex work is a contravention of morality spans histories, cultures and societies. This ubiquitous sentiment has spawned a widespread consideration of sex workers as deviants thereby providing a tacit impetus for violence against them (Pillai, Seshu, & Shivdas, 2008). The forms of violence suffered by FSWs include physical, sexual, economic, and psychological (Semple, Stockman, Pitpitan, & Strathdee, 2015; Fawole, & Dagunduro, 2014). Commonly associated risk factors include age, type of sex work practised (GAC, 2015), residential status of the FSW and educational status among others (Beksinska et al., 2018; George et al., 2011). The source of violence against FSWs are varied: the police or officers in uniform (Berger et al., 2018), clients, street thugs, family members, brothel owners/managers, and intimate partners are most commonly reported (GAC 2015; Schwitters et al., 2015 ; George et al., 2011). Besides, there are structural factors which contribute to violence against FSWs. In places where sex work is outlawed and punishable by law, FSWs suffer violence at the hands of law enforcement agents who come at them. Such prohibitive structures also deter FSWs from reporting the violence they suffer (Beyrer et al., 2015; Kate Shannon, Kerr, Strathdee, Montaner, & Tyndall, 2009; Toller et al., 2011). The narrative above is considerably representative of the dynamic presented on the Ghanaian landscape in terms of the risk factors for and sources of violence against FSWs. Across the ten Regions of Ghana, the 2015 IBBSS reports of varied degrees of 1 University of Ghana http://ugspace.ug.edu.gh violence against FSWs with roamers being more susceptible to forced sex than seaters. The study establishes the HIV prevalence among FSWs at 7%, which is more than four times that of the general population for the same year (1.62%). According to the 2014 Modes of Transmission (MoT) study, sex work accounted for 18.4% of total new HIV infections in Ghana (GAC, 2016). Globally, FSWs are highly vulnerable to and bear a disproportionately high portion of the HIV burden (K Shannon et al., 2016; Coetzee, Jewkes, & Gray, 2017; WHO, 2012; Beyrer et al., 2015). Being a venereal disease, HIV is transmitted mainly through unprotected sex with an infected person. Other risk factors include gender-based violence (Dunkle & Decker, 2013), age of sex debut (GAC, 2016), concurrent sexual partnership, serial monogamy, presence of other sexually transmitted infections, drug/substance use among others (WHO, 2018; Avert, 2018; AIDSinfo 2018; Boily et al., 2009). Inherent in sex work and FSW lifestyles are most of the risk factors mentioned above, thus heightening the vulnerability of sex workers to HIV acquisition and potential for HIV transmission. HIV remains a giant threat to public health globally (WHO, 2018). Since its advent, an estimated 77.3 million incidences have been recorded with 35.4 million deaths attributed to AIDS-related illnesses (UNAIDS, 2018). In 2017, an estimated 1.8 million new HIV infections were reported globally, 51% of which occurred among women (UNAIDS, 2016). Over two-thirds (69%) of the total estimates occurred in Sub-Saharan Africa against the background that Africa is home to only 11% of the world’s population. In 2011, 71% of all AIDS related deaths globally occurred in Sub-Saharan African (UNAIDS, 2012). The history of HIV in Ghana is traced to 1986 when the first forty-two cases were diagnosed (GAC, 2013). In 2017, the estimated adult national HIV prevalence was 2 University of Ghana http://ugspace.ug.edu.gh 1.67%, which translates into a national HIV population of 313,063. This comprised 91.0% (284,860) adults and 9.0% (28,203) children. The estimated adult HIV incidence rate in 2017 is 0.68% (19,101) with 15,694 deaths attributed to AIDS related causes (GAC, 2018). Key populations, including FSWs, and their sexual partners play a pivotal role in fuelling new HIV infections. In 2017, FSWs and their sexual partners contributed 47% of new HIV infections globally. It has also been established that FSWs have 13 times higher risk of HIV infection than the general population (UNAIDS, 2018). FSW and their sexual partners serve as a bridging population or vectors of transmission of HIV to the general population (Hesketh, Zhang, & Qiang, 2005; Santo Gomes do Espirito & Etheredge, 2005). The role of violence in heightening susceptibility to HIV acquisition and transmission has been severally established. As a consequence, violence against FSWs and women in general leads to inconsistent condoms use, condom failure, and other risks which exposes them to HIV acquisition (Decker, Lyons, et al., 2015; Platt et al., 2011b; Tounkara et al., 2014). This study shall identify the sources of violence against FSWs and establish its predictors as well as consequences with the goal to informing programming decisions among stakeholders of the national HIV response. For national HIV interventions to achieve holistic successes, it is necessary to design tailored interventions to reach KPs, including FSWs (GAC, 2016). This requires a detailed understanding of the milieu of sex workers and their modus operandi so that interventions will effectively respond to their needs. When FSWs are empowered with knowledge and tools, they can protect themselves and their clients from HIV and other preventable STIs. Although the IBBSS provides information on various aspects of sex work in Ghana including background characteristics of sex workers, it does not subject 3 University of Ghana http://ugspace.ug.edu.gh the data to analytical statistical modelling so that important details which miss the radar of descriptive statistics are lost. This study aims at using inferential statistical methods and tools to bring to bear unexplored details to facilitate advanced understanding of the data by key stakeholders. This is expected to provide further impetus to the country’s efforts at achieving the 90-90-90 targets and offer a sure footing toward ending AIDS by 2030 in line with the SDGs. 1.2 Statement of the problem In the last three months preceding the survey, the 2015 IBBSS reports that 17% of FSWs had been forced not to use condom during sexual intercourse and 10% were forced to have sex (non-consensual sex) against their will. These two incidences were peaked in the Central Region with 37% and 25% respectively. Also, in the last twelve months preceding the survey, 11% of FSWs reported having been physically assaulted by a client – Central Region leading with 29%; 4% offered sexual favours to avoid arrest by the police; and 9% reported having been beaten or physically abused by the police, clients, brothel owners and street thugs as a result of engaging sex work, with the highest incidence in Central Region (30%). Aside from being a violation of their fundamental human rights, violence against FSWs has inauspicious consequences for the national HIV response (Lyons et al., 2018; GAC, 2016). It has been severally established that violence against FSWs diminishes their ability to negotiate safer sex such as condom use thus making them more likely to engage in condomless sex which increases their risk of acquiring and transmitting HIV. It also leads to condom failure, promotes drug use among them as a coping mechanism, and leads to difficulty in reaching them with HIV services (Culture, August, & Taylor, 2010; Semple et al., 2015; Wirtz et al., 2015). 4 University of Ghana http://ugspace.ug.edu.gh HIV remains a giant threat to public health globally (WHO, 2018). Since its advent, an estimated 77.3 million incidences have been recorded with 35.4 million deaths attributed to AIDS-related illnesses (UNAIS, 2018). In 2017, an estimated 1.8 million new HIV infections were reported globally, 51% of which was fuelled by FSWs and their sexual partners (UNAIDS, 2018). The history of HIV in Ghana is traced to 1986 when the first forty-two cases were diagnosed. In 2017, the estimated adult national HIV prevalence was 1.67%, which translates into a national HIV population of 313,063 (GAC, 2018). HIV prevalence among FSWs as reported in the 2015 FSW IBBSS is 7%, which is more than four times the prevalence of the general population for the same year (1.62%). According to the 2014 Modes of Transmission (MoT) study, sex work solely accounted for 18.4% of total new HIV infections in Ghana (GAC, 2016). Evidently, the role of FSWs in reinforcing the HIV epidemic cannot be overemphasised. The activities of FSWs puts them at a greater risk of acquiring and transmitting HIV. By themselves and through their clients, FSWs serve as a bridge between the epidemic and the general populations. Consequently, reaching them with services is necessary to attain epidemic control toward ending AIDS (Jin et al., 2011; Etheredge, 2005). The IBBSS clearly establishes that FSWs in Ghana suffer violence. A number of factors were identified in the study as correlates of violence against FSWs and these include age, type of FSW and Region. Although other factors such as educational status of FSW, number of years in the sex industry, knowledge of violence and drug use have been identified in other studies (Semple et al., 2015; Fawole, & Dagunduro, 2014) as correlates of violence against FSWs, these were not explored in the IBBSS report even though data were collected on them. Besides, using only descriptive statistics, the IBBSS does not establish associations (if any) and degree of association between 5 University of Ghana http://ugspace.ug.edu.gh variables in order to ascertain their predictive powers. Again, the IBBSS does not establish any consequential relationship between violence and condom use, violence and condom failure, violence and HIV serostatus, violence and HIV testing behaviour among others. These and many others shall be investigated in this study. 1.3 Research questions 1. What are the sources of violence against female sex workers (FSWs) in Ghana? 2. What factors predict FSW vulnerability to violence in Ghana? 3. What are the consequences of violence against FSWs in Ghana? 1.4 Study objectives 1.4.1 General objective To assess the factors associated with violence against female sex workers in Ghana 1.4.2 Specific Objectives 1. To identify the sources of violence against FSWs in Ghana 2. To determine the predictors of FSWs’ vulnerability to violence in Ghana 3. To examine the consequences of violence against FSWs in Ghana 1.5 Justification This study uses analytical statistical modelling to unearth hitherto unknown details buried in the 2015 IBBSS in relation to violence against FSWs. Data gathered in the 2015 IBBSS were analysed mainly with the use of descriptive statistical tools. As noted in the conceptual and theoretical frameworks above, violence against FSWs is driven by a myriad of factors. Unfortunately, aside from age and type of FSW, the IBBSS does 6 University of Ghana http://ugspace.ug.edu.gh not address the rest even though data were collected on most of these variables. This places severe limitations on the applicability of the data for decision making. To address these limitations, it is necessary to subject the data to inferential statistical computations to determine association (if any) between the relevant variables. Moreover, it is needful to establish the levels of significance and predictive powers of the identified variables: both independently and collectively. Generally, descriptive statistics is considered the initial step toward further complicated modelling and analyses (Marsh & Stocker, 2010). Inherently, descriptive statistics is unable to bring to the fore nuanced and granular details to inform policy and programming decisions. Prevention has long been established as being better than cure, therefore it is expedient to have the ability to forecast FSWs’ vulnerability to violence in order to design appropriate interventions to pre-empt their occurrence. By employing analytical statistical modelling, this study seeks to provide answers to these questions. Additionally, Ghana has subscribed to the UNAIDS 90-90-90 fast track targets, and the Sustainable Development Goals. The tenets of the 90-90-90 fast track targets are that by 2020: • 90% of all persons living with HIV (PLHIV) should know their status • 90% of all who test HIV positive should be put and sustained on treatment • 90% of all who are on treatment should attain viral suppression state. Again, the SDG 3 states that countries should ensure healthy lives and promote well- being for all at all ages. One of the sub-sets of this goal is to end AIDS by 2030. Central to the achievement of these goals is the need to reach key populations, including FSWs with HIV prevention and other relevant services (GAC, 2016; USAID, 2017). This study seeks to shed light on one of the factors, if not the major factor, which hinders FSWs from accessing HIV services. When FSW are reached, new HIV 7 University of Ghana http://ugspace.ug.edu.gh infections will fall, AIDS related deaths will fall. This assertion is made on mounting evidence that when a person living with HIV (PLHIV) is put on sustained treatment, the viral load decreases until it reaches an undetectable state where effective transmission of the virus is eliminated (CDC, 2018). 1.6 Theoretical Framework: The Social Ecological Model This study takes its inspiration from the conceptual understanding of violence called the social ecological model, developed by Bronfenbrenner, which posits that partner violence is not the function of a single factor but the interplay of a complex web of factors at the individual, relationship, community and societal level (CDC, 2004; Campbell et al., 2009). Due to the high malleability of the model, it has been tailored to study a number of social phenomena including cyberbullying (Cross, 2015), partner violence (Heise, 2011), violence against women (Terry, 2014) and violence against FSWs (Platt et al., 2013). In the CDC’s adaptation of the model to study sexual violence and to model a prevention framework, it identifies societal level factors that increase the likelihood of violence to include religious or cultural belief systems, gender inequality, societal norms as well as economic or social policies that create or sustain gaps and tensions between groups of people. The community level factors include general tolerance for sexual assault, lack of assistance from police or judicial system, poverty, lack of employment opportunities, and weak community sanctions against perpetrators. At the relationship level, the following were identified: sexually aggressive peers or strongly patriarchal groups or persons. Alcohol and drug use, childhood history of sexual abuse, as well as harbouring attitudes and beliefs that support sexual violence were identified 8 University of Ghana http://ugspace.ug.edu.gh as the individual level factors that heightens one’s likelihood of perpetration or experiencing violence. Inspired by this conceptual model, this study shall consider the individual, occupational, structural and other factors that influence FSWs susceptibility to physical and sexual violence as detailed in the conceptual framework (figure 1). 1.7 Conceptual framework: Correlates of Violence FSWs’ vulnerability to violence Occupational - duration in sex work - length of stay in brothels - type of sex work practised - condom negotiation - price per sex - contract sex work - servicing clients in cars and public places - having clients who use drugs/alcohol Legal/structural Intermediary factors - anti sex work laws - inability to seek police - local policing practices protection - escaping to isolated and potentially dangerous Personal places Physical and - age of FSW - police brutality sexual violence - level of education - history of arrest against FSWs - drug/alcohol use - smoking - knowledge of violence - residency - financial insecurity Other - gender inequality - number of living children - societal anti sex work sentiments - being blackmailed - being a migrant - homelessness - having been trafficked into sex work Figure 1. Conceptual Framework: Risk factors for violence against FSWs Source: Based on literature review in this study. 9 University of Ghana http://ugspace.ug.edu.gh 1.7.1 Narrative on conceptual framework The conceptual framework above was developed based on evidence mobilised and crystallised from relevant literature. FSWs’ vulnerability to violence is identified with many factors. Work-related risk factors include duration in sex work (Wirtz et al., 2015); plying sex trade in brothels (Fawole & Dagunduro, 2014); having been trafficked into sex work; condom negotiation (Lyons et al., 2017; Schwitters et al., 2015); price per sex (Lima et al., 2017); being a migrant (Patel et al., 2016); being a contract sex worker; servicing clients in cars and public places violence (Shannon et al., 2009; Platt et al., 2013; Deering et al., 2014); and having clients who use drugs/alcohol, and financial insecurity (Reed et al., 2010). At the personal/individual level, the 2015 FSW IBBSS reported that FSWs who were above forty years were more vulnerable to forced sex than those who are younger. Other factors at the individual level which increases FSWs vulnerability to violence include years spent in sex work (Patel et al., 2016), low level of education, age, drug/alcohol use, smoking (Lima et al., 2017; Schwitters et al., 2015). However, FSWs’ vulnerability to violence is mitigated by the higher level of education (above second cycle and above) and knowledge of violence education (Schwitters et al., 2015). At the legal/structural level, it has been severally reported that the presence of anti-sex work laws and its attendant policing practices on sex workers expose them to violence. Anti-sex laws in and of themselves does not unleash violence on FSWs but unprofessional local policing practices and other intermediary factors serve as a nexus between these laws and violence against FSWs (Lyons et al., 2017; CDC 2004). Other factors that expose FSWs to violence are societal anti sex work sentiments, homelessness of FSW and gender inequality (Kate Shannon et al., 2009) . 10 University of Ghana http://ugspace.ug.edu.gh CHAPTER TWO LITERATURE REVIEW 2.0 Introduction In this chapter, literature relevant to the study are reviewed to put the study in context and to give the reader a panoramic view of the subject under investigation: violence against female sex workers. It begins with a summary of the 2015 IBBSS report which forms the bedrock of this study and continues to establish the determinants of violence against FSWs, as well as its consequences. 2.1 Summary of the 2015 IBBSS In response to one of the recommendations of the 2011 IBBSS, the GAC in collaboration with FHi360 and Noguchi Memorial Institute for Medical Research (NMIMR) carried out the 2015 IBBSS. This cross-sectional study employed a probability-based sampling approach to select survey participants (FSWs and their non- paying partners) from 56 Districts in all ten Regions of Ghana. Structured questionnaire and clinical procedures were used to collect behavioral data and biological data respectively which were later merged using STATA 12 software. Close to 90% of FSWs sampled were Ghanaians while Nigerians accounted for 6%. The remaining were other nationals, especially from other West African countries. The FSWs who participated in the study fell within 16-64 age brackets with a mean age of 26.8 (31.4years for seaters, and 25.5years for roamers). The mean age for initiation into sex work was 19.5 and 26.5 among roamers and seaters respectively. One-time paying clients, regular paying clients and non-paying partners are the three categories of sexual partners reported by the FSWs. The clients of FSW included drivers, artisans, the police and military and high school students. Also, the practice of 11 University of Ghana http://ugspace.ug.edu.gh having non-paying partners was common among the FSWs sampled with nearly 68% reporting to have at least one non-paying partner (Non-PP). In their most recent sex with a paying client prior the survey, about 90% of FSWs reported to have used a condom (seaters 96%; roamers 88%). However, this was not the case with their non-PPs in which only 21% reported consistent condoms use. In the last three months leading to the survey, about 10% of FSWs reported having been forced to have sex (either by a client or Non-PP). Alcohol and drug use among FSWs were also reported. In the last six months preceding the survey, 39% of FSWs reported having consumed alcoholic drink before sex while 17% had used marijuana. Other less reported substances include heroin, cocaine, sleeping pills and valium among others. Although about 72% of FSWs reported having ever tested for HIV, only 51% had the test in the past six months prior to the survey. In all, knowledge about condom use as an HIV preventive method was high (roamers, 78%; and seaters, 87%). HIV prevalence among FSWs in 2015 as reported in the survey was about 7% (roamers 5.4% and seaters 13.2 %) and that of other STIs are syphilis (less than 1%); Hepatitis B virus (roamers, 8.0%; seaters 7.7%); and Herpes Simplex Virus type 2 (67%). 2.2 Overview of sex work in Ghana Sex trade is plied in each of the ten Regions of Ghana. It is largely an urban phenomenon occurring mostly in Regional capitals, major border towns, busy market and transit towns and major mining towns. Majority of sex workers engage in sex work as their sole occupation and source of income, whereas others combine it with other engagements such as petty business, formal employment, agriculture, head pottering, and street hawking, maid service. Using the Multiplier Method, the 2015 Mapping and Population Size Estimation (MPSE) of FSWs in Ghana by the GAC estimated that there 12 University of Ghana http://ugspace.ug.edu.gh are 65,053 FSWs in Ghana with the highest proportion in Western (31.1%) and Greater Accra (16.7%) Regions. Upper East and Upper West Regions recorded the lowest proportions with 0.6% and 0.3% respectively. As reported in the 2015 IBBSS, majority (56%) of FSWs were introduced to sex work by friends with about one-third (35%) entering the sex trade by themselves. FSWs in Ghana, like other places are categorised into two groups: roamers and seaters. The GAC (2015), identified the following as places where FSWs take their clients for sex: home, hotel/guest house, bar, nightclub, lodge and other places. These other places as identified in the literature include the street, cars, sauna, parks, parking lots, or doorways (Schwitters et al., 2015; Platt et al., 2011; Shannon et al., 2009). As reported by GAC (2015), the mean age for commercial sex work debut was 21.0 years (95% CI: 20.34-21.60). Another study in Kumasi (Onyango et al., 2015) reported 12 years for commercial sex debut. Additionally, 40% of FSW reported to have been in sex work for between 2 and 5 years whereas a quarter had been in the trade for between 6 and 10years. Those who had been in the sex trade for over ten years was 14%. Most FSWs (80%) had never married before, only about 5% are currently married but 21% of seaters were divorced while 10% had separated (GAC, 2015). Majority (over 80%) of FSWs enter into sex work because of money for personal care, and to support their children and families. Yet, there are a few others who reported being forced into it by peer pressure. A study in Kumasi broadly classifies the motivations for entering into sex work into two: push and pull factors. The push factors include poverty, school dropout, migration, and unemployment. The pull factors on the other hand is summarized as peer pressure, that is mingling with friends who are involved in sex work, being envious of their affluence and eventually getting initiated into the trade (Onyango et al., 2015). As identified by Onyango et al., (2015), other 13 University of Ghana http://ugspace.ug.edu.gh studies in Soweto, South Africa also confirm that poor education outcome contributes to the drive to engage in sex trade (Coetzee et al., 2017). In Ghana, sex work is criminalised by the Ghana Criminal Code of 1960. Enforcing this law sometimes lead to a violation of the rights of FSWs at the hands of law enforcement agents. Violent arrests through police swoops and forced sex are commonly reported (Ansah, 2006; Onyango et al., 2015). According to the GAC (2016), the combined contribution of sex workers, partners of clients of sex workers, and clients of sex workers, is the second highest mode responsible for about 18% of all new HIV infections in Ghana. This makes it deserving of programmatic attention if the goal of the national HIV response is to be realised. 2.3 Gender-based violence and risk of HIV acquisition and/or transmission There is a gender dimension to violence against FSWs since most of the perpetrators are their clients and non-paying partners who happen to be males. Gender-based violence has long been established as a risk factor for HIV infection and transmission (Dunkle & Decker, 2013; Tounkara et al., 2014). In Togo and Burkina Faso, Andrea et al., (2015) reported that FSWs who experienced sexual violence are more likely to engage in condomless vaginal intercourse. Similar findings have also been reported in Mexico (Ulibarri et al., 2013), India (Beksinska et al., 2018), Nigeria (Fawole & Dagunduro, 2014), and South Africa (Coetzee et al., 2017). FSWs who experience violence are more likely to resort to drugs and other substance as a coping mechanism (Sarkar et al., 2005). These factors increase their vulnerability to HIV infection and transmission. As a result, FSWs are classified among the key populations in relation to HIV acquisition and transmission. 14 University of Ghana http://ugspace.ug.edu.gh 2.4 Sources and Types of Violence against Female Sex Workers Violence against FSW is a global phenomenon prevalent in every continent as demonstrated in the discourse below. In their study of violence against FSWs in two West African countries (Togo and Burkina Faso), Wirtz et al., (2015) established a self-reported lifetime physical abuse of 47.3%, forced sex (33.0%), and any violence (57.9%). The study mentions new clients, intimate and non-paying partners and strangers as the perpetrators of these acts. Another study in Abuja, Nigeria, (Fawole, & Dagunduro, 2014) established that over half of FSWs (52.5%) experienced violence against women (VAW) with the commonest being sexual violence (41.9%), followed by economic violence (37.7%), physical violence (35.7%) and psychological violence (31.9%). Physical violence was dominated by beating (78.4%) whereas unwelcome touch of breast or buttocks was the leading form of sexual violence (49.5%). Similar findings have been reported in the Gambia (Sherwood et al., 2015) which adds family members and uniformed police officers as perpetrators of physical violence. These findings are corroborated in other parts of Africa including Ghana (GAC 2015), Swaziland (Berger et al., 2018); Cote d’Ivoire (Lyons et al., 2017) and South Africa (Coetzee et al., 2017). In Latin America, Semple et al., (2015), in their study of violence against FSWs in thirteen Mexican cities established the prevalence of different types of client- perpetrated violence against FSWs: sexual (11.7%), physical (11.8%), economic (16.9%), and any violence (22.6%). This trend harmonises with the one reported by Wirtz et al., (2015). Higher prevalence of violence against FSWs have been reported in Brazil (Lima et al., (2017). 15 University of Ghana http://ugspace.ug.edu.gh Other studies in Asia, Northern America and Europe report various levels of violence against FSWs. High levels of sexual violence (77%) and physical violence (50%) have been reported in contract sex work in parts of India (George et al., 2011). In their systematic review and ecological analysis of factors mediating HIV risk among FSWs in Europe, Platt et al., (2013) established that there were consistently high incidences of violence against FSWs across the Region, and some of the studies reviewed estimated violence against FSWS to as high as 76% in Russia. The study further reported that in the last twelve months preceding the study, 16% of FSWs in Serbia (Kosova) reported being forced to have sex against their will. A lifetime experience of forced sex among street FSWs was 30% in Armenia, with 54% of street FSWs in Moldova (Eastern Europe) experiencing violence or been threatened by clients. 2.5 Determinants of FSWs susceptibility to violence FSWs’ vulnerability to violence is determined by many factors. For the purpose of this study, these factors shall broadly be compartmentalised into four: personal, occupational, legal/structural, and other. 2.5.1 Occupational factors that heighten FSWs’ vulnerability to violence Inherent in sex work are several factors that predispose FSWs to violence as discussed below. 2.5.1.1 Condom negotiation Most FSWs have knowledge of STIs including HIV and their modes of transmission, and are aware that correct and consistent use of condom can provide protection against HIV and other STIs (GAC, 2015). Consequently, some FSWs use condoms to prevent 16 University of Ghana http://ugspace.ug.edu.gh acquiring or transmitting HIV and other STIs. However, some clients of FSWs are not willing to use condom and this sometimes trigger violence from such clients when the FSW insist on using a condom. Increased frequency in clients demand for sex without condom has been associated with higher odds of experiencing physical abuse, verbal abuse, forced sex acts, being paid less, and not being paid at all (Lyons et al., 2017; Schwitters et al., 2015). The narrative above calls for greater empowerment of FSWs in condom negotiation self-efficacy because it has implications for their vulnerability to violence as established by Wirtz et al., (2015). This study estimated that over one-third (38.9%) of women who reported having suffered sexual violence indicated neutral to difficult condom negotiation self-efficacy with new clients compared with 19% of those who had experienced no sexual violence. 2.5.1.2 Servicing clients in open spaces The GAC (2015), identified the following as places where FSWs take their clients for sex: home, hotel/guest house, bar, nightclub, lodge and other places. These other places as identified in the literature include the street, cars, sauna, parks, parking lots, or doorways among others (Schwitters et al., 2015, 2015; Platt et al., 2011; Shannon et al., 2009). Places where FSWs service their clients has implications for their exposure to violence, especially client initiated violence (Shannon et al., 2009; Platt et al., 2013; Deering et al., 2014). Clients are less likely to initiate violence against an FSW if the sex occurred in the house of the FSW (aOR 0.12; 95 % CI: 0.03–0.49), the client’s house (aOR 0.17; 95 % CI: 0.04–0.74), or at a hotel (aOR 0.12; 95 % CI: 0.02–0.08) (Schwitters et al., 2015). This could be attributed to the herd security provided by the presence of other people at such places. Also, the client knowing that he might not go with impunity, is less likely to be violent. However, FSWs who service their clients in 17 University of Ghana http://ugspace.ug.edu.gh open spaces, such as those mentioned above, expose themselves to violence (Tara SH Beattie et al., 2010; Schwitters et al., 2015). FSWs who solicit for clients in public places or on phone suffer similar fate (Beksinska et al., 2018). 2.5.1.3 Servicing higher number of clients As the number of clients serviced by an FSW increases, her chances of meeting one or more who would be violent also increases (Tara SH Beattie et al., 2010). Beksinska et al., (2018) reported in their cross-sectional study of violence against FSWs in parts of India that FSWs who entertained an average of 9 clients per week were more likely to experience both workplace and community violence than those who entertained less. Such FSWs were also more likely to have started selling sex at a younger age (<25 years). In Karnataka State, South India, a mean of 12 clients per week (range 1-150) has been reported (Tara SH Beattie et al., 2010). Aside from having a relatively higher number of clients, Wirtz et al., (2015) further established that FSWs who had higher numbers of non-paying partners were also more likely to report incidence of forced sex. 2.5.1.4 Duration in sex work Even though experience of violence is commonly reported among FSWs who are new entrants to the trade (GAC, 2015; Beksinska et al., 2018), one’s vulnerability to violence also heightens with protracted continuity in the trade (Wirtz et al., 2015). Verbal abuse, forced sex, psychological violence, being paid less, and not being paid at all is commonly reported among FSWs who have been in the sex trade for relatively longer time, usually five years and above (Fawole & Dagunduro, 2014; Schwitters et al., 2015; Swain, Saggurti, Battala, Verma, & Jain, 2011; Wirtz et al., 2015). This is probably because as the FSW encounter new people all the time in their trade, it increases their probability of meeting someone who will subject them to one type of violence or another. Additionally, as an FSW enjoys a sustained period of violent-free 18 University of Ghana http://ugspace.ug.edu.gh run, she is likely to assume invincibility and thus engage in risky activities which will increase her susceptibility to violence. It is therefore necessary for programmers to continue to sensitise FSWS at all times on the need to avoid behaviours which will predispose them into becoming victims of violence. Other occupational factors which expose FSWs to violence include declining clients offer (Fawole & Dagunduro, 2014), and high cost per sex (Lima et al., 2017). 2.5.2 Personal Evidence abound in the literature about certain demographic and lifestyle characteristics of individual FSWs which heighten their vulnerability to violence. 2.5.2.1 Age Most FSWs experience one or multiple forms of violence during the initial stages of sex work (GAC, 2015; Beksinska et al., 2018). At this stage, most FSWs are below 20 years (GAC, 2015; Beksinska et al., 2018; Beattie et al., 2010) with very high susceptibility to sexual violence (GAC, 2015; Wirtz et al., 2015; (Silverman et al., 2011), physical violence, verbal violence (Lima et al., 2017) , and alcohol use (Beksinska et al., 2018; Lima et al., 2017; Schwitters et al., 2015). The GAC (2015) established that the reported incidence of forced sex decreases with the age of the FSWs, with a slight upsurge among FSWs of 40 years and above. This is confirmed by other studies in parts of India (Beksinska et al., 2018; Silverman et al., 2011); and Brazil (Lima et al., 2017). However, a contrary view is reported by Fawole, & Dagunduro., (2014) in a cross-sectional study at Abuja, Nigeria. This study reported significantly more incidence of violence (65.4%) among FSWs above 29 years than FSWs below 25 years (53.2%). Schwitters et al., (2015) in their studies did not find any association between the age of FSW and vulnerability to violence. 19 University of Ghana http://ugspace.ug.edu.gh 2.5.2.2 Knowledge and perception of violence among FSWs A study conducted in Abuja, Nigeria by Fawole & Dagunduro, (2014) established a hundred percent knowledge of VAW among FSWs. Regarding the forms of violence, beating was identified by majority (78.4%) as a form of physical violence; 49% stated unwelcome touch of buttocks and breasts as a form of sexual violence, with verbal abuse considered a form of psychological violence. Those who had attained post- secondary education had a highest knowledge of violence against women. Regarding their perception of VAW, the majority (73.1%) affirmed that VAW was prevalent in all professions. Consequently, over two-thirds (69%) accepted violence from clients as inherent in sex work so that it cannot be corrected. In the ensuing regression model, the study established that FSWs with poorer knowledge of VAW and a history of inter- parental violence were more susceptible to violence, especially physical violence. 2.5.2.3 Drug and alcohol use Schwitters et al., (2015) in their study of GBV among FSWs in Kampala, Uganda reported that the odds of experiencing client-initiated physical abuse were higher with FSWs who consumed five or more alcoholic drinks in the past six months preceding the study (aOR 2.60; 95 % CI: 1.54–4.43). In Brazil, Lima et al., (2017) established that alcohol and drug use predisposes FSWs to physical and verbal violence. The study further showed that drug use demonstrated the strongest association and statistical significance with physical and verbal violence from all the perpetrators studied. These findings resonate with the results of another study in Swaziland (Berger et al., 2018). A similar pattern has been reported for alcohol use. Unlike age, alcohol and drug use exhibits a strong association with all the types of violence suffered by FSWs. 20 University of Ghana http://ugspace.ug.edu.gh 2.5.2.4 Being HIV positive Of those who had experienced forced sex, 19% (13/40) were HIV positive compared to 15% (26/40) of women who had not experienced forced sex (Sherwood et al., 2015). Physical violence was associated with living with HIV in the crude model until adjusted for demographic characteristics (OR:1.8; aOR: 1.5; 95%CI: 0.8-2.8). (Lyons et al., 2017). 2.5.2.5 Financial insecurity Semple et al., (2015) established that FSWs reporting a “bad” or “extremely bad” financial footing had one and a half times (AOR = 1.51; 95% CI 1.04 – 2.18) higher odds of having suffered economic, sexual or physical violence. Being in debt is commonly reported among FSWs (Fawole & Dagunduro, 2014; Lyons et al., 2017; Patel et al., 2016; Reed et al, 2010). The work of Reed et al., (2010) sheds more light on the economic context of FSWs and its link to violence. This study, examining violence against FSWs in the context of economic insecurity, reports a very high percentage (80%) of borrowing among FSWs. The ensuing indebtedness and their inability to pay forces them into riskier sex trades with the goal to increasing earnings thus exposing them to violence. As reported by Reed et al., (2010), physical violence, unprotected sex, anal sex and STI were commonly pronounced among FSWs who reported debt. Indebtedness also influence FSWs decision to migrate (Patel et al., 2016) which further exposes them to violence. Some brothel owners also ask for free sex in exchange for FSWs inability to pay rent (Fawole & Dagunduro, 2014). 2.5.2.6 Residency (migrant or local dweller) In their study of the relationship between mobility and violence against FSWs, Patel et al., (2016) established a significantly higher experience of violence among FSWs who plied their trade outside their District than FSWs who had migrated to other places 21 University of Ghana http://ugspace.ug.edu.gh within their respective Districts. FSWs who had been in the sex trade for over five years had the highest tendency to migrate and ply their trade outside their political District. Mostly, such FSWs were above thirty years of age, solicited for clients using mobile phone and were currently or had ever been married. Migration was also reported by one-fifth of FSWs who reported being in debt. Beattie et al., (2010) affirm mobility of FSWs as a risk factor for physical and sexual violence. 2.5.2.7 Educational status In Abuja, Nigeria, Fawole & Dagunduro, (2014) established that the prevalence of violence among FSWs with no primary education is relatively higher than those with secondary and post-secondary education (Schwitters et al., 2015). The study further noted education provided some immunity to psychological and to some extent, economic violence, especially for FSWs who had received secondary or post-secondary education. This could be attributed to the empowerment which comes about through higher education. Besides, such women will have a better understanding of the nature of violence and how to navigate clearly off it. Educated women are also more likely to report threat and acts of violence against them thereby deterring potential perpetrators from targeting them. 2.5.3 Legal/Structural In a country like Ghana, where sex work is outlawed, criminalised and punishable by law, such legal underpinnings provide tacit approbation to violence against FSWs. 2.5.3.1 Law enforcement Local policing practices in countries where sex work is criminalised has consequence for violence against FSWs. There is widespread report of police rape or demand for sex of FSWs to avoid arrest or to be released from custody (Lyons et al., 2017; Rhodes et 22 University of Ghana http://ugspace.ug.edu.gh al., 2008). Violent arrest leading to physical violence is also a commonly reported incidence among FSWs in places where the practice is outlawed (GAC 2016). Police refusal to provide protection for FSWs who have experienced violence, police harassment, intimidation, and arrest among other hostile local policing practices have all been linked to sexual and physical violence (Lima et al., 2017; Lyons et al., 2017; Wirtz et al., 2015). FSWs are also forced by these policing practices to move away from main streets to isolated areas thereby exposing themselves to further violence (Shannon et al., 2009). 2.5.4 Other factors Gender inequality, societal anti-sex work sentiments, being blackmailed, homelessness and having been trafficked into sex work are factors which have been identified to contribute to FSWs vulnerability to violence. 2.5.4.1 Gender inequality Violence against FSWs is, by nature, an offshoot of violence against women, and most of the underlying factors fuelling the latter also apply to the former. Yodanis, (2004) conducted a cross-national test of the feminist theory of violence against women using data from the International Crime Victims Survey (ICVS) combined with United Nations statistics. His goal was to test the validity of feminist theory that the underlying factor for violence against women is gender inequality on the societal level. That is, men are more likely to be violent toward women the more unequal women are compared to men in a society. Using an adapted version of YLLӦ’s (1983, 1984) Status of Women Index, the study measured the status of women on the national level on three dimensions: educational, occupational, and political. The study concluded that the educational and occupational status of women in a country is inversely related to the 23 University of Ghana http://ugspace.ug.edu.gh prevalence of sexual violence against them. In other words, the higher the educational and occupational status of women in a country, the less likely men are to be sexually violent toward them. Specifically, it was established that 41% of variations in sexual violence against women were accounted for by the occupational status of women. Educational status on the other hand accounting for 40%. However, none of the three factors mentioned above was found to be significantly related to physical violence against women. In Tanzania, McCloskey et al., (2005) established that women with primary or lower education had a 1.7 higher odds of having been abused than their better educated peers. The study attributes the lower levels of violence suffered by more highly educated women to the empowerment education invariably confers on them such as more employment options which make them more valuable to their husbands. Again, such women are more likely to walk out of an abusive relationship because they have no fear of livelihood insecurities which confines many women to abusive partners. 2.5.4.2 Homelessness There exists a higher vulnerability to violence among FSWs who are homeless. After adjusting for potential confounding individual and interpersonal risk factors, Shannon et al., (2009) established that FSWs who were homeless had a 2.14 (95% CI 1.34 to 3.43) higher odds of being susceptible to violence. As established earlier, the odds of an FSW suffering a client-initiated violence is reduced if the client is serviced in the home of the FSWs. Conversely, there is increased odds of suffering client-initiated violence if the client is serviced in open places such as street, sauna, parks, parking lots, or doorways among others (Schwitters et al., 2015; Platt et al., 2011). Besides, FSWs who live on the street are more likely to be victims of rape, and violent police raid 24 University of Ghana http://ugspace.ug.edu.gh (Shannon et al., 2009). This calls for structural level interventions that addresses issues of poverty and affordable housing for the masses. 2.5.4.3 Societal anti sex work sentiments Ghana is predominantly a religious country. The 2010 Population and Housing Census established that 71.2% of Ghanaians profess the Christian faith, 17.6% Islam and 5.2% are adherents of the traditional religion (GSS, 2012). In the Old Testament times, the Bible prescribed capital punishment for sexual immorality defined as sex outside of marriage (Deuteronomy 22:13-21). Sex workers, referred to as harlots, prostitutes, or whores were to be eliminated by burning with fire (Genesis 38:24; Leviticus 21:9). In Islamic sexual jurisprudence, prostitution is outlawed (Quran 24:33) and it is commanded of prostitutes to be stoned to death. In the traditional Ghanaian society, there is no room for sexual perversion. Puberty rites such as Dipo among the Krobo and Bragoro among the Akan, as well as other practices such as female genital mutilation were all intended to discourage sexual promiscuity in the Ghanaian society. Teenage pregnancy, which is a visible outcome of sexual immorality, brings shame, not only to the people directly involved, but also to their families and the children born therefrom are pejoratively referred to as bastards. The foregoing narrative present the lens through which the wider Ghanaian society view sexual perversion. Consequently, the sentiment spawned by such socialisation cannot but be antagonistic toward sexual perversion. 2.5.4.4 Sex trafficking, contract sex work and related forms of sex work Female sex workers who were trafficked into sex work, those contracted into sex work and those who were forced into sex work have higher odds of susceptibility to violence than those who entered the trade of their own volition, and without any contractual bindings (Silverman et al., 2011). In Andhra Pradesh, India, George et al., (2011) 25 University of Ghana http://ugspace.ug.edu.gh established a 3.16 (95% CI, 2.01-4.95), and 2.14 (95% CI, 1.16-3.95) higher odds of susceptibility to physical violence among contract sex workers than non-contract sex workers respectively. Being threatened with a weapon, being threatened with beating, being actually hit, and being unable to control the number of clients served was more commonly reported among contract FSWs. High number of people being trafficked into sex work has been reported in Mumbai, India (Silverman et al., 2011). This study further established that FSWs reporting having been trafficked into sex work had significantly higher odds of sexual violence during their initial month in sex work (aOR 3.1; 95% CI, 1.6–6.1), a high volume of male clients (aOR, 3.3; 95% CI, 1.8–6.1), and no use of condoms in this same time period (aOR, 3.8; 95% CI, 2.1– 7.1). Sharing earnings with someone who provides a service related to sex work has also been associated with physical (OR:1.8; aOR:1.9; 95%CI:1.3-2.8 and sexual violence (OR:2.2; aOR:2.7; 95%CI: 1.8-4.1). Findings similar to the narrative above has been reported in Vancouver, Canada (Goldenberg et al., 2015). 2.5.4.5 Miscellaneous factors In addition to other factors, FSWs who had ever been married, are more vulnerable to violence than those who had never been married (Fawole & Dagunduro, 2014). Being blackmailed as a result of involvement in sex work was reported in Abidjan Côte d’Ivoire (Lyons et al., 2017; Sherwood et al., 2015), and Swaziland (Berger et al., 2018) and this incidence was associated with physical and sexual violence. Increased client demand for unprotected sex increased one’s odds for experiencing physical abuse six- fold, verbal abuse seven-fold; and five-fold each for forced sex acts; being paid less and not being paid at all (Schwitters et al., 2015). 26 University of Ghana http://ugspace.ug.edu.gh 2.6 Factors influencing HIV acquisition and transmission among FSWs As discussed below, FSWs’ vulnerability to HIV acquisition and transmission is determined by various of factors. 2.6.1 Multiple sexual partners Being a venereal disease, the risk of HIV infection is relatively higher among those with multiple sexual partners GAC (2016). By nature, sex workers establish sexual contacts with many people in the course of their work. In India, Beattie et al., (2010) reported a mean of 12 clients per week ranging from one to as high as one hundred and fifty. In Soweto, South Africa, Coetzee et al., (2017) reported that the majority (over 56.6%) of FSWs sold sex to 5–19 clients a day, and established that almost 60% of FSWs who serviced more than 5 clients per day were HIV positive. Another study in India (Bradley et al., 2011) confirms this and further adds that FSWs who had twenty or more clients were twice as likely to report condom breakage than those who had between 1 and 4. 2.6.2 High risk behaviour of clients Aside from FSWs engaging in sexual activities with multiple partners, their clients also exhibit similar behaviours by having multiple sexual partners including other FSWs. Patterson et al., (2009) in their study of risk behaviours among male clients of FSWs in parts of Mexico reported that aside from having a very high frequency of sexual contact with FSWs (25 times a year), and low condom use (50%), most of these clients (64%) reported that they never visited the same FSW twice. Other risk factors identified in the study include high drug use (88.3%), alcohol use (66.5%) and being high on drugs while having sex with an FSW (about 66%). A similar pattern of behaviour was reported of clients of FSWs in Yunnan, China by Jin et al., (2011). 27 University of Ghana http://ugspace.ug.edu.gh 2.6.3 Drug and alcohol use Drug and alcohol use have been identified as a risk factor for HIV transmission (Strathdee et al., 2015). This is because drugs and alcohol have the potential to impair one’s judgment and lead to involvement in high risk sex. Some FSWs also inject drugs leading to the sharing of needles which constitute an effective means of HIV transmission. Alcohol use has been associated with higher risk of condom breakage (Bradley et al., 2011). FSWs who are addicted to drugs may sell sex just to be able to buy the drugs required for daily injection. Such FSWs are also more likely to engage in unprotected sex, have higher number of clients and have sex with someone living with HIV (Platt et al., 2013). In Mumbai, India, Silverman et al., (2011) reported that women and girls who were trafficked into sex work used alcohol more frequently especially in the first month in sex trade and during their first commercial sex episode and were less likely to use condoms at such times. Violence against FSWs has also been established to increase the use of illicit drugs among them (Onyango et al., 2015). All these behavioural and lifestyle risks increase their vulnerability to HIV infection. 2.6.4 Experience of violence Violence against FSWs is a widespread phenomenon (Coetzee et al., 2017; Deering et al., 2014; Schwitters et al., 2015), and this exposes them to HIV acquisition in several ways. The form of violence which has a direct implication for HIV transmission is rape or forced sex, which unfortunately, is widely reported among FSWs (Fawole & Dagunduro, 2014; Schwitters et al., 2015; Tounkara et al., 2014). Forced sex is also associated with condom breakage (Bradley et al., 2011). In Ghana, the GAC (2015) established that 17% of FSWs were forced by their clients not to use condom during sex. Other forms of violence such as physical, economic, verbal and psychological have all been found to increase the victim’s risk of HIV acquisition. This is because violence 28 University of Ghana http://ugspace.ug.edu.gh has the potential to reduce the victim’s ability to negotiate for safer sex (Tounkara et al., 2014). It has also been established that men who are violent toward women are more likely to be HIV positive because they are more likely to engage in other forms of risky behaviours such as transactional sex, high-risk sex and drug use (Dunkle & Decker, 2013). In Abidjan, there is a widespread (80%) incidence of clients offering more money to FSWs to avoid using condom and this practice, aside from being a risk factor for HIV transmission, is also a form of sexual exploitation/violence (Lyons et al., 2017). 2.6.5 Presence of other STI STIs are commonly reported among FSWs (Le, D’Onofrio, & Rogers, 2010), and this constitute a risk factor for their vulnerability to HIV infection (GAC, 2016). In 2015, the GAC reported the following STIs among FSWs in Ghana: syphilis (1%); hepatitis B (8%); and Herpes Simplex virus type 2 (67%). In Soweto, South Africa, Coetzee et al., (2017) established that 50% of FSWs reported STI symptoms. A cross sectional study carried out in Cambodia demonstrated a strong relationship between Chlamydia trachomatis infection and HIV among FSWs and concluded that an active Chlamydia trachomatis infection is likely to increase the risk of HIV infection (Ohshige et al., 2000). Platt et al., (2013) in their systematic review uncovered reports of Chlamydia and gonorrhoea among FSWs in Europe. The narrative is a call to ensure that interventions targeting FSWs should be extended to cover any conceivable STIs because of their high vulnerability to STI infections. 2.6.6 Low condom use with non-paying partners and regular clients In Ghana, almost every FSW (99% of seaters and 98% of roamers) knows that correct and consistent use of condom is an HIV prevention method (GAC, 2015). This knowledge translates into consistent condom use among FSWs (92% of seaters; 75% of roamers) with paying clients. However, the narrative changes when it comes to non- 29 University of Ghana http://ugspace.ug.edu.gh paying partners. In Ghana the GAC (2015) reported that only 22% of FSWs consistently used condoms with their non-paying partners. A similar finding has been reported in Southern India where consistent condom use by FSWs with co-habiting partner or husband was 22.6% (Deering et al., 2011). In Adama, Ethopia, whereas there was almost 100% consistent condom use with non-regular clients, the same FSW population reported 38% inconsistent condom use with non-paying partners (Mooney et al., 2013). These episodes of low condom use with non-paying clients, coupled with the high-risk behaviour of these non-paying partners, exposes FSWs to HIV infection. 2.6.7 Early sex debut and duration in sex work Most FSWs enter the sex trade before the age of 20 (Berger et al., 2018; Silverman et al., 2011). At this stage their susceptibility to violence, including sexual violence is high (GAC, 2015; Beksinska et al., 2018). They are also more likely to report condom breakage (Bradley et al., 2011). Early sex work initiation coupled with protracted continuity in the trade has been established as a contributory factor for sexual networks and risk patterning (Shannon et al., 2015). 2.7 Consequence of violence against female sex workers Violence against FSWs is not without consequences. As discussed below, the consequence does not rest with the FSWs alone, it unfurls to cover the larger society. 2.7.1 Risk of HIV and other STI infections It has been variously established that violence against FSWs heightens their odds of acquiring HIV and other STI. FSWs who reported being beaten or raped in the past year were more likely to be infected with gonorrhoea compared with FSWs who did not (Tara SH Beattie et al., 2010). Experience of violence also limits their ability to negotiate safer sex/condom use (Tounkara et al., 2014) and further drives them to use 30 University of Ghana http://ugspace.ug.edu.gh drugs which is associated with HIV infection in many places including Taiwan (Lin et al., 2013). It also leads to increased tendency to migrate which exposes them to further violence including rape (Patel et al., 2016). FSWs are deterred from carrying paraphernalia such as condoms to evade police violence as such items may give them away as sex workers (Avert, 2018). Concerns about HIV risks are often overshadowed by fear of violence and exploitation and their predisposition to engaging in unprotected sex is heightened by these competing vulnerabilities (Onyango et al., 2015). 2.7.2 Increased migration As FSWs experience more violence in a particular geographical area, they seek safer havens by migrating to other places. Unfortunately, increased migration only serves to expose them to further risk of violence and these mutually reinforcing factors put them in an endless spiral of violence and migration. In places where sex work is criminalised, sex workers are unable to report the violence they suffer to police for fear of further violence (Decker, Crago, et al., 2015) and are therefore forced to abandon places of heightened violence to seek solace elsewhere. Also police raid and violent arrest further forces FSWs to resort to isolated places which further increases their vulnerability to violence (Decker et al., 2015; Shannon et al., 2009). 2.7.3 Negative mental health outcomes Depression is one of the most commonly reported mental health problems faced by FSWs, especially those who are victims of violence (Patel et al., 2016; Sherwood et al., 2015). Patel et al., (2016) established three-fold higher odds of screening positive for major depression among FSWs who reported being beaten or raped at least once in the past 1 year than those without such experience (AOR 3.0, 95% CI 2.4 to 3.6). Accounting for mobility makes the picture even worse. That is, FSWs who worked outside their district and reported violence were six times more likely to be at risk for 31 University of Ghana http://ugspace.ug.edu.gh major depressive symptoms (aOR 6.1, 95% CI 4.4 to 8.6) than FSWs who were not mobile and did not experience violence. Similarly, FSWs who were both mobile within the district and reported violence were three times more likely to be at risk for major depressive symptoms (43% vs 19%, aOR 3.0, 95% CI 2.2 to 4.2) than FSWs who were not mobile and did not experience violence. The odds of screening positive for major depression were seven times higher among FSWs who reported any violence perpetuated by the police (aOR 7.4, 95% CI 4.5 to 13.0), and five times higher among FSWs reporting violence perpetuated by a broker, pimp or goons (aOR 5.0, 95% CI 3.4 to 7.6) as compared with those with no history of having experienced violence. Other studies have established a relationship between violence against FSW and fear (Lyons et al., 2017) and suicidality (Berger et al., 2018). 2.7.4 Adverse Reproductive Health Outcome The prevalence of HIV and negative health outcome has been shown to be higher among FSWs who have experienced either physical or sexual violence (Swain et al., 2011). The same study further reported that FSWs with a history of having experienced violence of any sort had higher number of pregnancies, forced termination of pregnancy or multiple forced termination of pregnancies as compared to FSWs who had no history of having experienced violence. The study noted that historical and ongoing inconsistent condom use, symptoms of STI and a high perception of HIV risk was reported more among women who had ever experienced violence than those who had not experienced any violence. 2.7.5 Miscellaneous consequences Violence against FSWs results in loss of self-esteem, physical injury, unwanted pregnancy, and getting infected with sexually transmitted infections (Fawole & Dagunduro, 2014; Sherwood et al., 2015). Others include alcohol abuse (Mooney et al., 32 University of Ghana http://ugspace.ug.edu.gh 2013) which could be a coping mechanism due to police refusal to provide protection (Lyons et al., 2017). Fear of seeking health care as a result of engaging in sex work; and fear of walking in public places (Berger et al., 2018). Sex workers are also targets of serial killers and victims of homicide in many countries including Canada, the USA, Iran, Namibia, and the UK (Decker, Crago, et al., 2015). 2.8 HIV as a public health threat HIV remains a giant threat to public health globally (WHO, 2018). Since its advent, an estimated 77.3 million incidences have been reported with 35.4 million deaths attributed to AIDS-related illnesses (UNAIDS, 2018). In 2017, an estimated 1.8 million new HIV infections were reported globally, 51% of which occurred among women (UNAIDS, 2016). Over two-thirds (69%) of the total estimates occurred in Sub-Saharan Africa against the background that Africa is home to only 11% of the world’s population. In 2011, 71% of all AIDS related deaths globally occurred in Sub-Saharan African (UNAIDS, 2012). The history of HIV in Ghana is traced to 1986 when the first forty-two cases were diagnosed (GAC, 2013). In 2017, the estimated adult national HIV prevalence was 1.67%, which translates into a national HIV population of 313,063. This comprises 91.0% (284,860) adults and 9.0% (28,203) children. The estimated adult HIV incidence rate in 2017 is 0.68% (19,101) with 15,694 deaths attributed to AIDS related causes (GAC, 2018). Key populations, including FSWs, and their sexual partners play a pivotal role in fuelling new HIV infections (Patterson et al., 2009; Santo Gomes do Espirito & Etheredge, 2005). In 2017, FSWs and their sexual partners contributed 47% of new HIV infections globally. It has also been established that FSWs have 13 times higher 33 University of Ghana http://ugspace.ug.edu.gh risk of HIV infection than the general population (UNAIDS, 2018). FSW and their sexual partners serve as a bridging population or vectors of transmission of HIV to the general population (Hesketh et al., 2005; Santo Gomes do Espirito & Etheredge, 2005). 2.9 Conclusion The literature demonstrates that violence against FSWs is pervasive and their susceptibility to violence is influenced by a complex web of factors. Violence against FSWs have direct consequences on individual FSWs as well as rippling effects on society especially the risk of fuelling HIV infections among the general population. Therefore, it is in the interest of society that governments, CSOs and other social actors address this menace. 34 University of Ghana http://ugspace.ug.edu.gh CHAPTER THREE METHOD 3.0 Introduction Beginning with the study design, this chapter presents a description of the method for this study. It also provides information on the study site, variables, data analysis techniques and ethical considerations. 3.1 Study Design This study used secondary data from the 2015 Integrated Bio-Behavioural Sentinel Survey (IBBSS) among FSWs in Ghana which was a cross-sectional survey using quantitative data collection techniques. The study was conducted by the GAC and was split into two: Mapping and Size Estimation (MPSE) of FSWs and Bio-Behavioural Surveillance Survey (BBSS.) In the MPSE which preceded the BBSS, a snowballing approach was used to identify 536 hotspots, and using the multiplier method, an estimated 65,053 FSW was obtained (GAC 2015). For one to be included in the BBSS, the FSW had to be at least 16 years (because that is the age for consensual sex in Ghana). The operational definition of FSW was “females aged 16 years or older who reported to have received, in the last 6 months, money or favour as a source of income, from someone other than her established partner, in exchange for sexual services, and who may or may not consciously define these activities as ‘sex work’. Consequently, only 4751 FSWs were included in the BBS. However, after cleaning the data, this study used 4279 FSWs. The IBBSS distinguished two sub-types of FSW for the context of Ghana which are seaters and roamers. 35 University of Ghana http://ugspace.ug.edu.gh 3.2 Study Location The study was conducted in Ghana covering fifty-six (56) Districts in all ten Regions of the country. The study sites included FSW hotspots, all ten Regional capitals, all major border towns with FSW hotspots, popular market or transit towns, major mining towns and other relevant locations identified in the Mapping and Population Size Estimation (MPSE) study which preceded the bio-behavioural survey. 3.2.1 Re-demarcation of Regions in Ghana It is imperative to notify the reader that four Regions in Ghana have been re-demarcated after the 2015 IBBSS was conducted. Following a successful referendum in December 2018, six new Regions were created out of four existing Regions thereby increasing the number of Regions in Ghana from 10 to 16. The six new Regions are Savannah, North East, Bono East, Ahafo, Oti and Western North. Also, former Brong Ahafo Region was re-named Bono Region following the re-demarcation. Savannah and North East were carved out of Northern Region; Ahafo and Bono East, out of former Brong Ahafo; Oti and Western North out of Volta and Western Regions respectively. Consequently, Ghana, as used in this study refers to Ghana before the 2018 referendum. Below are two maps indicating the old and new Ghana maps. 36 University of Ghana http://ugspace.ug.edu.gh Figure 2. Map of Ghana (before 2018 Figure 3. Map of Ghana (after the 2018 referendum) Referendum) Source: http://www.ghana.gov.gh/index.php/about- Source: http://1d1f.gov.gh/ ghana/regions 3.3 Variables 3.3.1 Dependent Variables The dependent variables for this study are physical violence, and sexual violence. On physical violence, the question asked was “During the past 12 months, did any client use physical force or violence on you?” and the response was dichotomized into “1= yes; 0 = no.” Sexual violence on the other hand was measured with three questions: 1. “During the last 3 months, did any one – paying or otherwise – force you to have sex with him when you didn’t want to?” with dichotomized response options: “1= yes; 0 = no”. 2. “During last 3 months, have you been forced by any client not to use condoms during sex?” and the response was dichotomized into “1= yes; 0 = no.” 3. During the last 3 months, how many times have you avoided arrest by providing the police officer with a “sexual favor”? This question was recoded to “0 = all 37 University of Ghana http://ugspace.ug.edu.gh who provided no sexual favour to avoid arrest by police” and “1 = all who provided at least one sexual favour to avoid arrest by police.” Missing variables and non-responses were dropped. A new variable was created to represent sexual violence which combines all three questions measuring sexual violence. It was recoded “0 = for all who answered no to all three questions” and “1 = for all who answered yes to at least one of the three questions” 3.3.2 Independent Variables Based on evidence in the literature, many factors influence physical and sexual violence against FSWs (as detailed in figure 1. Conceptual Framework, page 6,). However, those variables relevant for this study and which could be answered with the data from the 2015 IBBSS among FSWs are the following: Socio-demographic factors: The relevant socio-demographic variables used in the study are age, educational status, marital status, number of living children, region, and nationality. Age started sex work (age at which FSW started sex work): This was measured by “At what age did you start sex work?” It was grouped into <20; 20-29; 30-39; 40+ and the “<20” category was used as the reference in the multiple logistic analysis due to literature that this category are prone to experiencing more violence (GAC, 2015; Wirtz et al., 2015; (Silverman et al., 2011). Alcohol use: Measured using questions such as “In the past 6 months, have you drunk alcoholic drinks before sex?” and “During the last 4 weeks, how often have you had drinks containing alcohol?” In the multiple logistic analysis, those who used alcohol 38 University of Ghana http://ugspace.ug.edu.gh were used as the reference because the literature shows they are more susceptible to violence (Lima et al., 2017). Substance use: This variable was measured by “Have you ever used any such substances [e.g. cocaine, heroin, weed, medicines such as valium & sleeping pills and any other substances that alter your mood that are not prescribed by a doctor] even once?” Those who have ever used substance were used as the reference group in the multiple logistic analysis due to their higher susceptibility to violence compared to those who do not (Lima et al., 2017). Type of FSW: this was measured by whether the FSW is a “roamer” or “seater” Other source of income: measured by “Other than sex work, do you have another means of earning money?” Where FSW takes clients for sex: measured by “Where do you usually take your clients for sex?” The response options were “1=Home; 2=Hotel/Guest House; 3=Bar; 4=Nightclub; 5=Lodge; 9=Other (specify); 99=no response” External financial support for children: measured by “Apart from your support, have you received other external financial support for the children in the last 12 months?” This is because FSWs who are in financial distress are more likely to engage in riskier behaviors that predisposes them to violence (Fawole & Dagunduro, 2014) therefore receiving some form of external financial support for children might serve to mitigate such risks. Contact with peer educators: measured by “Have you been in contact with any health peer educator in the community in the last 12 months?” 39 University of Ghana http://ugspace.ug.edu.gh 3.4 Study Population The study population comprised 4,279 FSWs in all ten Regions of Ghana. 3.5 Preparation of data for analysis The 2015 IBBSS dataset was received from the Ghana AIDS Commission as a STATA file. First of all, all missing values (denoted in STATA by a dot); “don’t know” (denoted by 98) and non-responses (denoted in the study by “99”) were removed from all independent and dependent variables. This reduced the sample size from the original 4,751 to 4,279. Secondly, there was recoding; grouping, regrouping and dichotomisation of responses: Age (current biological age) was grouped into 16-20; 21-30; 31-40; 40+; Age_2 (age at which FSW started sex work) was grouped into <20; 20-29; 30-39; 40+; region was re- ordered during the analysis in line with violence experiences instead following the alphabetical ordering style, same was done for nationality during the analysis. marital status was dichotomised into “currently married vs not currently married” (Decker, Crago, et al., 2015); Substance use: measured by “Have you ever used any such substances [e.g. cocaine, heroin, weed, medicines such as valium & sleeping pills and any other substances that alter your mood that are not prescribed by a doctor] even once?” was dichotomized into “0=never used any substance before; 1=has ever used at least one of the substances mentioned above or other substances.” Number of living children was grouped into 0; 1-2; 3+ (Beksinska et al., 2018). Where clients are taken for sex among the response options for this variable, lodge was merged with “Hotel/guest house” and Bar was merged with nightclub (due to their similarities and the smaller number of respondents on each of them) thus bringing the total response options to 4. 40 University of Ghana http://ugspace.ug.edu.gh 3.6 Bivariate Analysis Data was analysed using STATA 15 (StataCorp LLC). Pearson’s Chi-squared (χ2) was used to establish the association between the independent variables and the outcome variables. This test was used because the variables are categorical, and recorded at least 5 for expected value of the number of sample observations (Mchugh, 2013). Consequently, two tables were constructed: one for physical violence and another for sexual violence and how the various variables under consideration are associated with each of them. All variables with a p-value value above 0.20 were excluded from the logistic regression model except where their relevance have been established in the literature. 3.7 Simple and multiple logistic regression analysis Logistic regression was used to determine the direction and strength of association between the independent variables and the outcome variables. This is because the outcome is dichotomous (0=No; 1=Yes), and the variables are categorical and the goal is to model the outcome. Crude Odds Ratios were obtained from the simple logistic regression. Afterward the independent variables were adjusted to obtain the Adjusted Odds Ratios using the multiple logistic regression model. A p-value below 0.05 was used to declare statistical significance. 41 University of Ghana http://ugspace.ug.edu.gh 3.8 Ethical Consideration The Ghana AIDS Commission, collectors of the primary data, obtained ethical clearance for the study from the FHI 360’s Protection of Human Subject Committee (PHSC) and the Noguchi Memorial Institute for Medical Research (NMIMR) ethical review boards. Written consent was sought and obtained from the management of GAC to use the secondary data for this study. 42 University of Ghana http://ugspace.ug.edu.gh CHAPTER FOUR RESULTS 4.0 Introduction This chapter presents the results of the study based on the analysis of data. It begins with demographic information about FSWs and proceeds to a more rigorous analysis using inferential statistics. 4.1 Demographic Characteristics of FSWs The IBSS study involved a total of 4751 FSWs; however, after cleaning the data based on the criteria described in the Methods section of this study, 4,279 FSWs satisfied the inclusion criteria for this study. As indicated in Table 1, majority (91.31%) of the FSWs were Ghanaians with very few (6.16%) coming from countries that do not share geographical borders with Ghana (namely Nigeria and others). Majority (63.52%) of them were between the ages of 20 and 29 even though more than half (53.54%) of them started sex work before age 20 with a median duration of 4 years in sex work. Only a few (4.46%) of them reported that they were currently married, though over half (51.37%) of them reported having at least one living child. About a third (30.43%) had either primary or no formal education with only 3% having attained post-secondary education. 43 University of Ghana http://ugspace.ug.edu.gh Table 1. Demographic characteristics Variable Frequency Percentage (n=4,279) current age 16-19 581 13.58 20-29 2,718 63.52 30-39 752 17.57 40+ 228 5.33 Mean (sd, range) 26.17 (7.00; 16, 64) Marital status currently married 191 4.46 not currently married 4,088 95.54 Number of living children 0 2,081 48.63 1-2 1,778 41.55 3+ 420 9.82 Median (1st quartile, 3rd quartile; range) 1 (0, 1; 0, 8) Age started sex work <20 2,291 53.54 20-24 1,198 28.00 25-29 441 10.31 30+ 349 8.16 Median (1st quartile, 3rd quartile; range) 19 (17, 23; 10, 62) Duration in sex work <5 years 2,647 61.86 5+ years 1,632 38.14 Median (1st quartile, 3rd quartile; range) 4 (2,8; <1, 47) Educational Status No formal education 566 13.23 Primary 736 17.20 Middle 102 2.38 JSS 1,747 40.83 Secondary/SSS/Vocation 998 23.32 Higher 130 3.04 Region Ashanti 452 10.56 Brong Ahafo 434 10.14 Central 402 9.39 Greater Accra 842 19.68 Eastern 400 9.35 Northern 278 6.50 Upper East 386 9.02 Upper West 175 4.09 Western 513 11.99 Volta 397 9.28 Nationality Ghana 3,907 91.31 Nigeria 227 5.30 Togo 54 1.26 Ivory Coast 11 0.26 Burkina Faso 43 1.00 Other 37 0.86 44 University of Ghana http://ugspace.ug.edu.gh 4.2 Sources of violence against FSWs This section begins with a brief description of the forms of physical and sexual violence experienced by FSWs in Ghana as contained in the 2015 IBBSS among FSWs and proceeds with analysis of the sources of these forms of violence. In all, 433 (10.12%) of the FSWs reported having suffered physical violence in the past 12 months preceding the survey. The same number reported having experienced forced sex in the past three months while 155 (3.26%) reported having provided sexual favours to the police to avoid arrest. The reported prevalence of forced sex without condom in the past three months preceding the survey was 17.71%. The topmost reasons cited by respondents for inconsistent condom use with paying clients are client refusal and fear of client becoming violent (23.91%). These same reasons were also cited among others for inconsistent condom use with non-paying partners. Another form of sexual violence/exploitation reported in the study is clients paying more to avoid using condoms 143 (18.33%). As indicated in figure 4 more than half (50.58 %) of the reported incidence of physical violence against FSWs was perpetrated by paying one-time clients; about a third (30.95%) by paying regular clients; and the rest by spouses, lovers and boyfriends. Of the perpetrators of forced sex, 189 (43.65%) were paying one-time client, 149 (34.41%) were paying regular clients, and 124 (28.64%) being spouses, lovers and boyfriends. Per the study, these one-time paying clients and regular paying clients comprised long distance truck drivers, artisans, police/military, High School students, hand cart pushers, foreigners and others. 45 University of Ghana http://ugspace.ug.edu.gh one-time paying clients regular paying clients other s e x u a l P REVALENCE P HYSICAL SEXUAL TYPE AND SOURCE OF VIOLENCE Figure 4. Prevalence, type and sources of violence against FSWs 4.3 Bivariate analysis of factors associated with physical violence against FSWs This study established 10.12% prevalence of physical violence against FSWs. From Table 2, the following demographic features of FSWs are associated with their experience of physical violence: current age of FSW (χ2=10.69; p=0.01); marital status (χ2=5.24; p=0.02); Region of residence (χ2=257.83; p<0.01); and age at which FSW started sex work (χ2=9.43; p=0.02). Other demographic characteristics such as education, nationality and number of living children were not associated with physical violence against FSWs. Other variables which had association with experience of physical violence among FSWs are type of FSW (χ2=9.45; p<0.01); where clients are taken for sex (χ2=8.75; p=0.03); whether FSW has ever used drugs (χ2=50.20; p<0.01); whether an FSW ever drunk alcohol before sex in the past six months (χ2=28.66; p<0.01) and whether an FSW has other source(s) of income besides sex work (χ2=17.97; p=0.02). 46 PERCENT 10 25 51 31 28 44 34 29 University of Ghana http://ugspace.ug.edu.gh Table 2: Bivariate analysis of factors associated with physical violence against FSWs n=4,279 Ever suffered physical violence p-value Chi-value n (%) No Yes Type of FSW <0.01 9.45 Roamer 3383 (79.06) 3,016 (89.15) 367 (10.85) Seater 896 (20.940 830 (92.63) 66 (7.37) Current age 0.01 10.69 16-19 581 (13.58) 515 (88.64) 66 (11.36) 20-29 2718 (63.52) 2,437 (89.66) 2 81 (10.34) 30-39 752 (17.57) 675 (89.76) 77 (10.24) 40+ 228 (5.33) 219 (96.05) 9 (3.95) Education 0.12 8.77 None 566 (13.23) 522 (92.23) 44 (7.77) Primary 736 (17.20) 648 (88.04) 88 (11.96) Middle 102 (2.38) 96 (94.12) 6 (5.88) JSS 1747 (40.83) 1562 (89.41) 185 (10.59) SSS/vocational 998 (23.32) 901 (90.28) 97 (9.72) Higher 130 (3.04) 117 (90) 13 (10) Marital status 0.02 5.24 not currently married 4,088 (95.54) 3665 (89.65) 423 (10.35) currently married 191 (4.46) 181 (94.76) 10 (5.24) Region <0.01 257.83 Ashanti 452 (10.56) 398 (88.05) 54 (11.95) Brong Ahafo 434 (10.14) 399 (91.94) 35 (8.06) Central Region 402 (9.39) 285 (70.90) 117 (29.1) Eastern 400 (9.35) 376 (94.00) 24 (6) Greater Accra 842 (19.68) 728 (86.46) 114 (13.54) Northern 278 (6.5) 242 (87.05) 36 (12.95) Upper East 386 (9.02) 369 (95.60) 17 (4.4) Upper West 175 (4.09) 167 (95.43) 8 (4.57) Volta Region 397 (9.28) 375 (94.46) 22 (5.54) Western 513 (11.99) 507 (98.83) 6 (1.17) Nationality 0.23 6.88 Burkina Faso 43 (1) 39 (90.70) 4 (9.3) Ghana 3907 (91.31) 3,519 (90.07) 388 (9.93) Ivory Coast 11 (0.26) 9 (81.82) 2 (18.18) Nigeria 227 (5.30) 203 (89.43) 24 (10.57) Togo 54 (1.26) 47 (87.04) 29 7 (12.96) Other 37 (0.86) (78.38) 8 (21.62) Age started sex work 0.02 9.43 <20 2291 (53.54) 2,031 (88.65) 2 6 0 (11.35) 20-29 1639 (38.3) 1,491 (90.97) 1 4 8 (9.03) 30-39 282 (6.59) 261 (92.55) 21 (7.45) 40+ 67 (1.57) 63 (94.03) 4 (5.97) Duration in sex work 0.22 1.53 <5 years 2647 (61.86) 2,391 (90.33) 2 5 6 (9.67) 5+ years 1632 (38.14) 1,455 (89.15) 1 77 (10.85) No. of living children 0.10 4.55 0 2081 (48.63) 1,891 (90.87) 190 (9.13) 1-2 1778 (41.55) 1,579 (88.81) 199 (11.19) 3+ 420 (9.82) 376 (89.52) 44 (10.48) 47 University of Ghana http://ugspace.ug.edu.gh Table 2: Bivariate analysis of factors associated with physical violence against FSWs (continued) n=4,279 Ever suffered physical violence p-value Chi-value n (%) No Yes Where clients are 0.03 8.75 taken for sex Home 1702 (39.78) 145 (8.52) 1,557 (91.4) Hotel/guest house/ 2408 (56.27) 272 (11.30) 2,136 (88.70) lodge Bar/nightclub 73 (1.71) 6 (8.22) 67 (91.78) Other 96 (2.24) 10 (10.42) 86 (89.58) Ever used drug <0.01 50.20 No 3,224 (75.34) 2,958 (91.750 266 (8.25) Yes 1,055 (24.66) 888 (84.17) 167 (15.83) Drunk alcohol before <0.01 28.66 sex (past 6 months) No 2548 (59.55) 2,342 (91.92) 206 (8.08) Yes 1731 (40.45) 1,504 (86.89) 227 (13.11) Other source of income 0.02 17.97 None (No other job) 2861 (66.86) 2,594 (90.67) 267 (9.33) Formal employment 114 (2.66) 106 (92.98) 8 (7.02) Petty business 737 (17.22) 643 (87.25) 94 (12.75) Agriculture 10 (0.23) 9 (90) 1 (10) Head potter 45 (1.05) 37 (82.22) 8 (17.78) Maid servant 12 (0.28) 12 (100) 0 (0.00) Bar attendant 61 (1.43) 57 (93.44) 4 (6.56) Street hawker 129 (3.01) 119 (92.25) 10 (7.75) Other 310 (7.24) 269 (86.77) 41 (13.23) 4.4 Bivariate analysis of factors associated with sexual violence against FSWs Forced sex without condom was the highest reported form of sexual violence against FSWs with a prevalence of 17.71% followed by forced sex (10.12%), and provision of sexual favours to evade arrest by police (3.26%). Sexual violence as used in this study is the aggregation of all who reported having suffered at least one of the above forms of sexual violence, and the prevalence of sexual violence as established by this study is 24.75%. From Table 3, the following demographic factors were associated with sexual violence: current age of FSW, measured by age at last birthday, (χ2=17.53; p=0.01); level of education (χ2=12.87; p=0.03); marital status (χ2=14.59; p<0.01); region of residence (χ2=246.23; p<0.01); and nationality (χ2=13.67; p=0.02). Other factors associated with FSWs experience of sexual violence are where clients are taken for sex (χ2= 13.95; 48 University of Ghana http://ugspace.ug.edu.gh p<0.01); whether an FSW has ever used drug(s) (χ2= 20.36); whether an FSW has ever drunk alcohol before sex in the past six months (χ2=18.50; p<0.01); and whether an FSW has other source(s) of income (χ2=28.85; p<0.01). However, number of living children (χ2=5.91; 0.05), type of FSW (χ2=3.59; p=0.06), age at which FSW started sex work (χ2=8.36; p=0.08; and duration in sex work (χ2=8.36; p=0.31) were not associated with sexual violence. Table 3. Bivariate analysis of factors associated with sexual violence against FSWs Ever suffered Sexual violence n (%) No Yes p-value Chi-value Type of FSW 0.06 3.59 Roamer 3383 (79.06) 2,524 (74.61) 8 59 (25.39) Seater 896 (20.94) 696 (77.68) 200 (22.32) Current age 0.01 17.53 16-19 581 (13.58) 425 (73.15) 1 5 6 (26.85) 20-29 2718 (63.52) 2,016 (74.17) 7 0 2 (25.83) 30-39 752 (17.57) 585 (77.79) 1 6 7 (22.21) 40+ 228 (5.33) 194 (85.09) 3 4 (14.91) Education 0.03 12.87 None 566 (13.23) 442 (78.09) 124 (21.91) Primary 736 (17.20) 552 (75.00) 184 (25.00) Middle 102 (2.380 86 (84.31) 16 (15.69) JSS 1747 (40.83) 1,277 (73.10) 4 70 (26.90) SSS/vocational 998 (23.32) 760 (76.15) 238 (23.85) Higher 130 (3.04) 103 (79.23) 27 (20.77) Marital status <0.01 14.59 not currently married 4088 (95.54) 3,054 (74.71) 1,034 (25.29) currently married 191 (4.46) 166 (86.91) 25 (13.09) Region <0.01 246.23 Ashanti 452 (10.56) 356 (78.76) 96 (21.24) Brong Ahafo 434 (10.14) 361 (83.18) 73 (16.82) Central Region 402 (9.39) 194 (48.26) 208 (51.74) Eastern 400 (9.35) 327 (81.75) 73 (18.25) Greater Accra 842 (19.68) 639 (75.89) 203 (24.11) Northern 278 (6.50 229 (82.37) 49 (17.63) Upper East 386 (9.02) 344 (89.12) 42 (10.88) Upper West 175 (4.09) 129 (73.71) 46 (26.29) Volta Region 397 (9.28) 291 (73.30) 106 (26.70) Western 513 (11.99) 350 (68.23) 163 (31.77) 49 University of Ghana http://ugspace.ug.edu.gh Table 3. Bivariate analysis of factors associated with sexual violence against FSWs (continued) Ever suffered Sexual violence n (%) No Yes p-value Chi-value Nationality 0.02 13.67 Burkina Faso 43 (1) 36 (83.72) 7 (16.28) Ghana 3907 (91.31) 2,929 74.97) 978 (25.03) Ivory Coast 11 (0.26) 9 (81.82) 2 (18.18) Nigeria 227 (5.3) 185 (81.50) 42 (18.50) Togo 54 (1.26) 40 (74.07) 14 (25.93) Other 37 (0.86) 21 (56.76) 16 (43.24) Age started sex work 0.08 6.66 <20 2291 (53.54) 1,715 (74.86) 5 7 6 (25.14) 20-29 1639 (38.3) 1,223 (74.62) 4 1 6 (25.38) 30-39 282 (6.59) 226 (80.14) 5 6 (19.86) 40+ 67 (1.57) 56 (83.58) 1 1 (16.42) Duration of sex work 0.31 1.03 <5 years 2647 (61.86) 2,391 (90.33) 2 5 6 (9.67) 5+ years 1632 (38.14) 1,455 (89.15) 1 7 7 (10.85) No. of living children 0.05 5.91 0 2081 (48.63) 482 (23.16) 1,599 (76.84) 1-2 1778 (41.55) 472 (26.55) 1,306 (73.45) 3+ 420 (9.82) 105 (25.00) 315 (75.00) Where clients are taken <0.01 13.95 for sex Home 1702 (39.78) 1,288 (75.68) 414 (24.32) Hotel/guest house/ 2408 (56.27) 1,788 (74.25) 6 20 (25.75) lodge Bar/nightclub 73 (1.71) 57 (78.08) 16 (21.92) Other 96 (2.24) 87 (90.63) 9 (9.38) Ever used drug <0.01 20.36 No 3224 (75.34) 2,481 (76.95) 7 43 (23.05) Yes 1055 (24.66) 739 (70.05) 316 (29.95) Drunk alcohol before sex <0.01 18.50 (past 6 months) No 1731 (40.45) 1,243 (71.81) 488 (28.19) Yes 2548 (59.55) 1,977 (77.59) 5 71 (22.41) Other source of income <0.01 28.85 None (No other job) 2861 (66.86) 2,139 (74.76) 7 22 (25.24) Formal employment 114 (2.66) 98 (85.96) 16 (14.04) Petty business 737 (17.22) 578 (78.43) 159 (21.57) Agriculture 10 (0.23) 7 (70.00) 3 (30.00) Head potter 45 (1.05) 33 (73.33) 12 (26.67) Maid servant 12 (0.28) 9 (75.00) 3 (25.00) Bar attendant 61 (1.43) 48 (78.69) 13 (21.31) Street hawker 129 (3.01) 104 (80.62) 25 (19.38) Other 310 (7.24) 204 (65.81) 106 (34.19) 50 University of Ghana http://ugspace.ug.edu.gh 4.5 Predictors of physical violence against FSWs From the study, the following variables showed a statistically strong independent association with physical violence against FSWs: type of FSW (p-value<0.01), current age of FSW (p-value=0.02), Region (p-value<0.01), whether FSW has ever used drugs (<0.01) and whether FSW had ever used alcohol before sex in the past six months preceding the survey (p-value<0.01). Compared with roamers, seaters have 47% (aOR 0.53; 95% CI: 0.37 – 0.75) diminished odds of experiencing physical violence whilst FSWs who are above 40 years have 79% (aOR 0.21; 95% CI: 0.07 – 0.63) less odds of experiencing physical violence than FSWs currently below twenty years. FSWs who did not use alcohol before sex in the past six months have 31% (aOR 0.69; 95% CI: 0.56 – 0.86) reduced odds of experiencing physical violence. Other variables such as education, marital status, nationality, age at which FSW started sex work, duration in sex work, number of living children, where clients are taken for sex and whether FSW has other source of income were not independently significant after the adjustment as detailed in Table 4. Table 4. Multiple logistic regression analysis of factors influencing physical violence against FSWs Unadjusted 95% CI Adjusted 95% CI p-value OR Type of FSW <0.01 Roamer Ref Ref Ref Ref Seater 0.65 0.50 – 0.86 0.53 0.37 – 0.75 Current age 0.02 16-19 Ref Ref Ref Ref 20-29 0.90 0.68 – 1.20 0.75 0.52 – 1.07 30-39 0.89 0.63 – 1.26 0.83 0.50 – 1.38 40+ 0.32 0.16 – 0.65 0.21 0.07 – 0.63 Education 0.32 None Ref Ref Ref Ref Primary 1.61 1.10 – 2.36 1.50 1.00 – 2.26 Middle 0.74 0.31 – 1.79 1.08 0.42 – 2.77 JSS 1.41 1.00 – 1.98 1.16 0.80 – 1.69 SSS/vocational 1.28 0.88 – 1.85 1.18 0.79 – 1.77 Higher 1.32 0.69 – 2.53 1.64 0.81 – 3.34 Marital status 0.15 not currently married Ref Ref Ref Ref currently married 0.48 0.25 – 0.91 0.60 0.30 – 1.20 51 University of Ghana http://ugspace.ug.edu.gh Table 4. Multiple logistic regression analysis of factors influencing physical violence against FSWs (continued) Unadjusted 95% CI Adjusted 95% CI p-value OR Region <0.01 Upper East 0.11 0.07 – 0.19 0.11 0.06 – 0.20 Brong Ahafo 0.21 0.14 – 0.32 0.24 0.16 – 0.37 Northern 0.36 0.24 – 0.55 0.43 0.27 – 0.68 Eastern 0.16 0.10 – 0.25 0.15 0.10 – 0.25 Ashanti 0.33 0.23 – 0.47 0.38 0.26 – 0.56 Greater Accra 0.38 0.28 – 0.51 0.57 0.40 – 0.80 Volta Region 0.14 0.09 – 0.23 0.15 0.09 – 0.24 Upper West 0.12 0.06 – 0.24 0.14 0.06 – 0.29 Western 0.03 0.01 – 0.07 0.04 0.02 – 0.09 Central Region Ref Ref Ref Ref Nationality 0.82 Nigeria Ref Ref Ref Ref Burkina Faso 0.87 0.29 – 2.64 1.58 0.47 – 5.27 Ivory Coast 1.88 0.38 – 9.21 2.03 0.37 – 11.25 Ghana 0.93 0.60 – 1.44 0.97 0.60 – 1.59 Togo 1.26 0.51 – 3.10 1.41 0.53 – 3.70 Other 2.33 0.96 – 5.68 0.86 0.32 – 2.32 Age started sex work 0.22 <20 Ref Ref Ref Ref 20-29 -0.25 -0.47 – -0.04 0.79 0.60 – 1.04 30-39 -0.46 -0.93 – 0.00 0.69 0.37 – 1.29 40+ -0.70 -1.72 – 0.32 1.64 0.38 – 7.13 Duration in sex work 0.62 <5 years 0.88 0.72 – 1.08 0.93 0.71 – 1 .22 5+ years Ref Ref No. of living children 0.36 0 0.859 0.607 – 1.214 0.74 0.48 – 1.14 1-2 1.077 0.76 – 1.521 0.83 0.56 – 1.25 3+ Ref Ref Ref Ref Where clients are taken 0.63 for sex Hotel/guest house/ Ref Ref Ref Ref lodge Home 0.73 0.59 – 0.90 1.10 0.86 – 1.42 Bar/nightclub 0.70 0.30 – 1.64 0.68 0.28 – 1.66 Other 0.91 0.47 – 1.78 0.85 0.42 – 1.75 Ever used drug <0.01 No 0.48 0.39 – 0.59 0.56 0.45 – 0.70 Yes Ref Ref Ref Ref Drunk alcohol before <0.01 sex (past 6 months) No 0.58 0.48 – 0.71 0.69 0.56 – 0.86 Yes Ref Ref Ref Ref Other source of income 0.22 None (No other job) Ref Ref Ref Ref Formal employment 0.73 0.35-1.52 0.37 0.44 – 2.05 Petty business 1.42 1.11-1.82 0.19 1.04 – 1.80 Agriculture 1.08 0.14-8.55 0.99 0.10 – 7.83 Head potter 2.10 0.97-4.56 0.88 0.94 – 4.79 Maid servant Bar attendant 0.68 0.25-1.89 0.36 0.23 – 1.94 Street hawker 0.82 0.42-1.58 0.32 0.44 – 1.78 Other 1.48 1.04-2.11 0.25 0.85 – 1.84 Adjusted for all variables in the table p-value obtained from Wald test Reference category maid servant is empty from the model because it predicted the failure perfectly 52 University of Ghana http://ugspace.ug.edu.gh 4.6 Predictors of sexual violence against FSWs As presented in Table 5, the following variables proved independently significant in predicting sexual violence against FSWs after controlling for all other potential confounders: current age of FSW (p-value=0.02), marital status (p-value<0.01), number of living children (p-value=0.04), where FSWs take clients for sex (p-value <0.01), Region (p-value<0.01), whether FSW has a history of drug use (p-value <0.01), and whether FSW has ever used alcohol before sex in the past six months (p- value<0.01). However, differences exist in the odds of susceptibility to violence in the individual units of these predictive variables. For example, compared with FSWs who are below 20 years, FSWs who are above 40 years are 62% less likely to experience sexual violence (aOR 0.38; 95% CI: 0.21–0.71). In the same vein, FSWs who reported no history of drug use have 20% lower odds of experiencing sexual violence than those who reported having ever used drugs (aOR 0.80; 95% CI: 0.68 – 0.95). Other variables such as type of FSW, education, nationality, age at which FSWs started sex work, duration in sex work and whether FSW has other source of income than sex work was not independently statistically significant after adjusting for all other confounders. 53 University of Ghana http://ugspace.ug.edu.gh Table 5. Multiple logistic regression analysis of factors influencing sexual violence against FSWs Crude OR 95% CI Adjusted OR 95% CI p-value Type of FSW 0.14 Roamer Ref Ref Ref Ref Seater -0.17 -0.34 – 0.01 0.84 0.67 – 1.06 Current age 0.02 16-20 Ref Ref Ref Ref 21-30 -0.05 -0.26 – 0.15 0.85 0.67 – 1.10 31-40 -0.25 -0.50 – 0.00 0.67 0.47 – 0.96 40+ -0.74 -1.15 – -0.33 0.38 0.21 – 0.71 Education 0.96 None Ref Ref Ref Ref Primary 1.61 1.10 – 2.36 1.07 0.82 – 1.41 Middle 0.74 0.31 – 1.79 0.86 0.46 – 1.59 JSS 1.41 1.00 – 1.98 1.09 0.85 – 1.39 SSS/vocational 1.28 0.88 – 1.85 1.07 0.82 – 1.40 Higher 1.32 0.69 – 2.53 1.03 0.63 – 1.70 Marital status <0.01 not currently married Ref Ref Ref Ref currently married -0.81 -1.24 – -0.38 0.42 0.27 – 0.67 Region <0.01 Upper East -2.17 -2.55 – -1.80 0.12 0.08 – 0.18 Brong Ahafo -1.67 -1.99 – -1.35 0.19 0.14 – 0.27 Northern -1.61 -1.98 – -1.25 0.24 0.16 – 0.35 Eastern -1.57 -1.89 – -1.25 0.22 0.16 – 0.31 Ashanti -1.38 -1.68 – -1.08 0.27 0.20 – 0.36 Greater Accra -1.22 -1.47 – -0.96 0.37 0.28 – 0.49 Volta Region -1.08 -1.38 – -0.78 0.44 0.32 – 0.60 Upper West -1.10 -1.49 – -0.71 0.38 0.25 – 0.57 Western -0.83 -1.10 – -0.56 0.48 0.36 – 0.64 Central Region Ref Ref Ref Ref Nationality 0.71 Nigeria Ref Ref Ref Ref Burkina Faso -1.37 -2.40 – -0.33 1.47 0.58 – 3.74 Ivory Coast -1.21 -1.94 – -0.48 0.65 0.13 – 3.36 Ghana -1.23 -2.90 – 0.43 1.31 0.90 – 1.91 Togo -0.82 -1.48 – -0.17 1.39 0.67 – 2.90 Other -0.78 -1.67 – 0.11 1.22 0.55 – 2.74 Age started sex work 0.88 <20 Ref Ref Ref Ref 20-29 0.01 -0.13 – 0.16 1.08 0.89 – 1.31 30-39 -0.30 -0.61 – 0.00 1.04 0.68 – 1.59 40+ -0.54 -1.19 – 0.12 1.10 0.46 – 2.65 Duration in sex work <5 years 1.14 0.93 – 1 .39 1.02 0.85 – 1.23 5+ years Ref Ref Ref Ref No. of living children 0.04 0 0.86 0.61 – 1.21 0.69 0.51 – 0.93 1-2 1.08 0.76 – 1.52 0.80 0.60 – 1.06 3+ Ref Ref Ref Ref 54 University of Ghana http://ugspace.ug.edu.gh Table 5. Multiple logistic regression analysis of factors influencing sexual violence against FSWs (continued) Crude OR 95% CI Adjusted OR 95% CI p-value Where clients are <0.01 taken for sex Ref Hotel/guest house/ Ref Ref Ref lodge 0.92 – 1.30 Home 0.73 0.59 – 0.90 1.09 0.50 – 1.62 Bar/nightclub 0.70 0.30 – 1.64 0.90 0.14 – 0.60 Other 0.91 0.47 – 1.78 0.29 Ever used drug <0.01 No 0.48 0.389 – 0.588 0.80 0.68 0.95 Yes Ref Ref Ref Ref Drunk alcohol before <0.01 sex (past 6 months) No -0.31 -0.45 – -0.17 0.74 0.64 – 0.86 Yes Ref Ref Ref Ref Other source of 0.01 income Ref Ref Ref Ref None (No other job) 0.48 0.28-0.83 0.54 0.31 – 0.94 Formal employment 0.81 0.67-0.99 0.77 0.63 – 0.95 Petty business 1.27 0.33-4.92 1.20 0.29 – 4.92 Agriculture 1.08 0.55-2.10 1.16 0.58 – 2.34 Head potter 0.99 0.27-3.66 0.92 0.23 – 3.63 Maid servant 0.80 0.43-1.49 0.82 0.42 – 1.57 Bar attendant 0.71 0.46-1.11 0.72 0.45 – 1.16 Street hawker 1.54 1.20-1.98 1.41 1.07 – 1.86 Other Adjusted for all variables in the table; p-value obtained from Wald test 4.7 Consequences of violence against FSW This section looks at the effects of physical and sexual violence against FSWs on three variables: condom use, HIV testing behaviour and HIV test outcome. Condom use is sub-divided into consistent condom use and condom failure (condom break/tear/burst during sexual activity). The bivariate analysis using chi-squared revealed that there is association between physical violence and consistent condom use (p-value<0.01), and condom failure (p- value<0.01). The same nexus exists between sexual violence and the aforementioned two variables. However, neither sexual violence nor physical violence had any relationship with HIV testing behaviour and HIV test outcome. 55 University of Ghana http://ugspace.ug.edu.gh Compared with FSWs who had ever experienced physical violence, FSWs who had never experienced physical violence have 58% higher odds of using condoms consistently with paying clients (OR 1.58; 95% CI 1.26 – 1.98). Also, FSWs who reported no experience of sexual violence, were 49% less likely to report condom failure (OR 0.51; 95% CI 0.44 – 0.59). Although not statistically significant, FSWs who reported no history of having experienced sexual violence were 10% more likely to access HTS than those who had ever suffered sexual violence (OR 1.10; 95% CI 0.94 – 1.30). Details are presented in the tables below (table 6a to table 6d) Table 6. Analysis of violence against FSWS and consistent condom use with paying clients Consistent condom use (n= 4,250) No n (%) Yes n (%) Unadjusted 95%CI χ2 p- OR value Ever suffered <0.01 physical violence No 759 (19.88) 3,058 (80.12) 1.58 1.26 – 1.98 Yes 122 (28.18) 3 1 1 ( 7 1 . 8 2 ) Ref Ref Ever suffered <0.01 sexual violence No 598 (18.72) 2 , 5 96 (81.28) 1.59 1.35 – 1.87 Yes 283 (26.80) 7 7 3 (73.20) Ref Ref Table 7. Analysis of violence against FSWs and condom failure Condom failure (n= 4,108) No n (%) Yes n (%) OR 95%CI χ2 p-value Ever suffered <0.01 physical violence No 2,659 (72.14) 1,027 (27.86) 0.60 0.49 – 0.74 Yes 257 (60.90) 165 (39.10) Ref Ref Ever suffered <0.01 sexual violence No 2,307 (74.59) 786 (25.41) 0.51 0.44 – 0.59 Yes 609 (60.00) 406 (40.00) Ref Ref 56 University of Ghana http://ugspace.ug.edu.gh Table 8. Analysis of violence against FSWS and HIV testing behaviour HIV test (past 6 months) (n= 4,108) No n (%) Yes n (%) OR 95%CI χ2 p-value Ever suffered 0.42 physical violence No 1,294 (49.15) 1,339 (50.85) ** ** Yes 166 (51.55) 156 (48.45) Ever suffered 0.23 sexual violence No 1,075 (48.77) 1,129 (51.23) 1.10 0.94 – 1.30 Yes 385 (51.26) 366 (48.74) Ref ** not included in regression analysis because χ2 p-value >0.2 Table 9. Analysis of violence against FSWS and HIV test outcome HIV test outcome (n= 3,493) Negative n (%) Positive n (%) OR 95%CI χ2 p-value Ever suffered 0.98 physical violence No 2,916 (93.64) 198 (6.36) ** ** Yes 355 (93.67) 24 (6.33) Ever suffered 0.71 sexual violence No 2,423 (93.55) 167 (6.45) ** ** Yes 848 (93.91) 55 (6.09) 57 University of Ghana http://ugspace.ug.edu.gh CHAPTER FIVE DISCUSSION 5.0 Introduction This chapter discusses the results of the study presented in the previous chapter and attempts to relate the findings to the literature. 5.1 Socio-demographic characteristics Even though sex work is outlawed in Ghana, it is a highly thriving urban enterprise with a predominantly youth and young adult population. As part of the study protocol, the IBBSS did not consider FSWs below 16 years, but this study noted that the sex industry in Ghana runs the gamut of extremely young population to overly adult population as some FSWs reported having started sex work at an early age of 10 years whilst some continue in the outlawed trade even at 62 years of age. Consistent with findings from other studies (Decker, Lyons, et al., 2015; Fawole & Dagunduro, 2014), the vast majority (over 90%) of FSWs were not married even though over 50% of them reported having at least one child. Fawole & Dagunduro, (2014) attribute the lack of interest in marriage among FSWs to the perceived and actual restrictions marriage will pose to their trade. The paradox of low marriage against many having given birth raises a question on consistency of condom use and other sexually risky behaviours among FSWs in Ghana and other places. As reported in other studies (Mooney et al., 2013; Onyango et al., 2015; Platt et al., 2011b), the sex trade in Ghana is no respecter of one’s education or employment status: even persons with post-senior high education and formal employment were engaged in the trade even though they constituted the minority. This places a call on government and society to create opportunities for girl-child education and women empowerment 58 University of Ghana http://ugspace.ug.edu.gh because highly educated women and employed women are less likely to engage in sex work. 5.2 Sources of violence against FSWs As confirmed by the literature (Tara SH Beattie et al., 2010; Silverman et al., 2011; Wirtz et al., 2015), the major sources of violence against FSWs in Ghana are (one- time/new) paying clients, regular paying clients and non-paying partners Client perpetrated violence is often triggered by client refusal to use condoms, price of sex or the use of alcohol and/or drugs by either the client or the FSW. Besides, FSWs are also surrounded by a complex web of gatekeepers including non-paying partners, facility (bar, nightclub, hotel/motel etc.) owners who regulate the lives of these FSWs, and all these actors have been linked with violence against FSWs (Tara SH Beattie et al., 2010). The IBBSS did not explore the details of these complex actors so this study is limited in its ability to analyse the role of these actors in perpetuating violence against FSWs. The specific source which the IBBSS touches is the role of law enforcement agents (the police) who sometimes capitalise on the illegality of the trade to threaten arrest and detention in order to gain sexual favours from the FSWs. This is a widely reported incidence in places where sex work is outlawed and this episode is also linked with physical and sexual violence against FSWs (Tara SH Beattie et al., 2010; Tounkara et al., 2014). There is the need for comprehensive programmatic interventions to address these structural contributions to violence against FSWs because there is always a dilemma between proper enforcement of the law against the illegality of the trade and the encroachment on the rights of the perpetrators. 59 University of Ghana http://ugspace.ug.edu.gh 5.3 Predictors of violence against FSWs In FSWs’ experience of physical and sexual violence, region of residence was a significant predictor of each of them. Central region recorded the highest in terms of both physical (29.10%) and sexual (51.74%) violence and the difference between central Region and any other Region was statistically significant. Aside from region of residence, each of these variables also predicted physical and sexual violence against FSWs: current age of FSW, whether FSW has ever used drugs (lifetime), and whether FSW had ever used alcohol before sex in the past six months preceding the survey. In relation to current age, FSWs who were below 20 years were more susceptible to both physical and sexual violence than those who were above 20 years and the odds of experiencing either physical or sexual violence was least among FSWs who were over 40 years of age. For example, compared with FSWs below 20 years, FSWs above 40 years had 79% and 62% less odds of experiencing physical and sexual violence respectively. These predictors have been validated by other studies (Tara SH Beattie et al., 2010; Lima et al., 2017; Wirtz et al., 2015). Type of FSW (roamer or seater) could only predict FSWs experience of physical violence (p-value<0.01) with seaters having 47% less odds of experiencing physical violence than roamers. This is because roamers are prone to scouting for clients in unsafe places that exposes them more to physical violence, and they are also more susceptible victims of police raids which is associated with both physical and sexual violence. Marital status (p-value<0.01), number of living children (p-value=0.04), where FSWs take clients for sex, and whether FSW has an alternative source of income (p- value=0.01) were only predictors of sexual violence but not physical violence. 60 University of Ghana http://ugspace.ug.edu.gh Consistent with other studies (Decker, Lyons, et al., 2015), FSWs who were married were less likely to experience sexual violence than those who are not. Also, FSWs who had who had children were more susceptible to sexual violence than those who are not, and those who had one or two children were less susceptible to experiencing sexual violence than those who had three or more living children. This is because the extra financial burden posed by the care of their children forces them into riskier activities in their desperation to make extra money to meet their expenses (Semple et al., 2015). In the same vein FSWs who were formally employed had the least susceptibility to violence in comparison with FSWs who had no other source of income aside from sex work as well as FSWs who were engaged in menial or other work besides sex work. The IBBSS mentions forced sex, forced sex without condom and using sexual favours to evade arrest by police as the types of sexual violence experienced by FSWs. It does not mention the types of physical violence experienced by FSWs, but relevant examples from the Domestic Violence in Ghana report include slapping, throwing things at victim, choking or strangling as well attacking with a weapon. Other types of sexual violence mentioned in the report include sexual comments and using intimidation to elicit sex. All these dimensions require further investigation and redress. 5.4 Consequences of violence against FSWs As established in Kumasi (Onyango et al., 2015) and elsewhere (Tara S Beattie et al., 2016; Mooney et al., 2013), this study noted a significant association between FSWs experience of violence and consistency of condom. That is, FSWs who experienced either physical or sexual violence or both were less likely to use condoms consistently than those who had no history of having experienced violence. Consistent condom use among FSWs is important in controlling new HIV infections in the general population 61 University of Ghana http://ugspace.ug.edu.gh because the clients of FSWs serve as vectors of HIV transmission to the general population (Jin et al., 2011; Etheredge, 2005). Therefore, the national response to HIV should explore the nexus between violence and condom use and carve a strategy to address the menace. Like Coetzee et al. (2017), this study did not establish any significant association between FSWs experience of violence and their HIV positive sero-status. This could be due to the low HIV prevalence among the general population in Ghana. However, the evidence for experience of violence and HIV positivity has been established in other studies (Dunkle & Decker, 2013; Dunkle et al., 2017; Tounkara et al., 2014) and this caveat is for consideration by HIV programmers among FSWs. Neither was experience of violence associated with HIV testing behaviour. As confirmed in other studies (Bradley et al., 2011; Decker, Lyons, et al., 2015; Tounkara et al., 2014), this study established a strong association between experience of violence and condom failure (condom breakage, tearing or bursting). Condom failure has the potential to increase the risk of HIV infection and this requires programmatic attention. Lessons could be drawn from the findings of Bradley et al. (2011) in their study of determinants of condom breakage among female sex workers in parts of India. They established that women who received condoms from peer educators and women who had ever seen condom demonstration were less likely to experience condom breakage. This is an encouragement to programmers to strengthen their education and sensitisation drive as it holds promise of addressing the incidence of condom breakage. 62 University of Ghana http://ugspace.ug.edu.gh CHAPTER SIX CONCLUSIONS AND RECOMMEDATIONS 6.0 Introduction This chapter presents the conclusions of the study and makes recommendations based on the findings and conclusions. 6.1 Conclusions The conclusions below follow from the objectives and findings of the study. The study established that violence against FSWs is prevalent in Ghana, and the following sources of violence against FSWs were identified: clients, non-paying partners, the police, brothel owners, and street thugs. The following variables independently predict FSWs experience of either physical or sexual violence in Ghana: Region of residence, current age of FSW, whether FSW has ever used drugs (lifetime), and whether FSW had ever used alcohol before sex in the past six months preceding the survey. Marital status, number of living children, where FSWs take clients for sex, and whether FSW has an alternative source of income were independent predictors of sexual violent but not physical violence, whereas type of FSW predicted only physical violence but not sexual violence. Consequently, there is a strong evidence of association between experience of physical or sexual violence among FSWs and consistent condom use as well as condom failure. 6.2 Recommendations Based on the findings and conclusions of the study, the following are recommended: 1. Like any other GBV in Ghana, there is widespread condonation of violence against FSWs in the Ghanaian society. Therefore, any attempt at addressing the 63 University of Ghana http://ugspace.ug.edu.gh menace must first tackle the underlying socio-cultural and structural factors that fuel these sentiments. 2. Strategies to address violence against FSWs should have tailored and concentrated interventions for younger FSWs, FSWs with children, FSWs who are not currently married, FSWs in Central Region as well as FSWs with low education and no/menial employment. 3. In pursing the first 90 (getting at least 90% of all PLHIV to know their HIV status) of the 90-90-90 targets, stakeholders of the national HIV response should treat violence against FSWs as a human right issue, but not a strategy for mobilising them for testing since the two are independent of each other. 4. There is urgent need to address the factors (identified by Onyango et al., 2015) that drive minors into sex work through comprehensive social interventions and the creation of safety nets for vulnerable and destitute children. Also, there is the need to address the broader macro-economic factors that plunge women and girls into sex work through education and the creation of viable means of livelihood because the commonest drive for their engaging in sex work is the quest to secure a livelihood. 5. There is the need for comprehensive sexuality education and the inculcation of values in young girls through strong social institutions like religion, culture or the family so that female adolescents and youth could be socialised to make healthy sexual and reproductive health choices from their early years. 6. HIV programmers should strengthen condom negotiation and condom use skills of FSWs to address the incidences of low condom use and condom failure among FSWs with a history of having experienced violence. 64 University of Ghana http://ugspace.ug.edu.gh 6.3 Limitations 1. In a country like Ghana where sex work and drug use are outlawed, people are likely to offer only politically correct (or socially desirable answers) to questions on such matters especially those which they deem incriminating. 2. Recall bias. The study assumed that respondents were sincere with their answers to the research questions. So, where a respondent, either inadvertently or otherwise reported inaccurate information, the credibility of this study will be limited by the extent to which that information was inaccurate. Also, judging from the fact that many of the FSWs reported having ever used alcohol or drugs or both, it is possible that some of the respondents gave answers under the influence of these substances which tend to impair one’s judgment. 3. Personal and social perception of violence. A society’s cultural, social and political definition of violence creates a context through its members perceive and interpret their experience of violence. Such pervading subjectivity has the potential to introduce biases in reporting one’s experience of violence. 65 University of Ghana http://ugspace.ug.edu.gh REFERENCES AIDSinfo. (2018). HIV/AIDS: The Basics. https://aidsinfo.nih.gov/understanding-hiv- aids/fact-sheets/19/45/hiv-aids--the-basics Date accessed: 18.01.2019 Ansah, N. (2006). Structural Relations of the Sex Trade and Its Link to Trafficking : The Case of Ghana Author ( s ): Nancy Ansah Source : Agenda : Empowering Women for Gender Equity , No . 70 , Gender-Based Violence Stable URL : http://www.jstor.org/stable/4066739 Structu. Empowering Women for Gender Equity, 1(70), 100–106. https://doi.org/10.2307/4066739 Avert (2018) How do you get HIV https://www.avert.org/hiv-transmission-prevention/how- you-get-hiv 18.01.2019 Beattie, Tara S, Isac, S., Bhattacharjee, P., Javalkar, P., Davey, C., Raghavendra, T., … Heise, L. (2016). Reducing violence and increasing condom use in the intimate partnerships of female sex workers : study protocol for Samvedana Plus , a cluster randomised controlled trial in Karnataka state , south India. BMC Public Health, 15–17. https://doi.org/10.1186/s12889-016-3356-7 Beattie, Tara SH, Bhattacharjee, P., Ramesh, B., Gurnani, V., Anthony, J., Isac, S., … Moses, S. (2010). Violence against female sex workers in Karnataka state, south India: impact on health, and reductions in violence following an intervention program. BMC Public Health, 10(1), 476. https://doi.org/10.1186/1471-2458-10-476 Beksinska, A., Prakash, R., Isac, S., Mohan, H. L., Platt, L., Blanchard, J., … Beattie, T. S. (2018). Violence experience by perpetrator and associations with HIV / STI risk and infection: a cross-sectional study among female sex workers in Karnataka, south India. BMJ Open, 6–10. https://doi.org/10.1136/bmjopen-2017-021389 Berger, B. O., Grosso, A., Adams, D., Ketende, S., Sithole, B., Mabuza, X. S., … Baral, S. (2018). The Prevalence and Correlates of Physical and Sexual Violence Affecting Female Sex Workers in Swaziland. Journal of Interpersonal Violence, 33(17), 2745 – 2766. https://doi.org/10.1177/0886260516629385 Beyrer, C., Crago, A. L., Bekker, L. G., Butler, J., Shannon, K., Kerrigan, D., … Strathdee, S. A. (2015). An action agenda for HIV and sex workers. The Lancet. https://doi.org/10.1016/S0140-6736(14)60933-8 Boily, M., Baggaley, R. F., Wang, L., Masse, B., White, R. G., Hayes, R. J., & Alary, M. (2009). Heterosexual risk of HIV-1 infection per sexual act : systematic review and meta-analysis of observational studies. The Lancet Infectious Diseases, 9(February). https://doi.org/10.1016/S1473-3099(09)70021-0 Bradley, J., Rajaram, S., Alary, M., Isac, S., Washington, R., Moses, S., & Ramesh, B. M. (2011). Determinants of condom breakage among female sex workers in Karnataka , India. BMC Public Health, 11(Suppl 6), S14. https://doi.org/10.1186/1471-2458-11-S6- S14 Campbell, R., Dworkin, E., & Cabral, G. (2009). An Ecological Model of the Impact of Sexual Assault On Women’s Mental Health. Trauma, Violence, & Abuse, 10(3), 225– 246. https://doi.org/10.1177/1524838009334456 Centers for Disease Control and Prevention (2004). Sexual violence prevention: beginning the dialogue. Atlanta, GA: Centers for Disease Control and Prevention; 2004. Centers for Disease Control and Prevention (2018). HIV Treatment as Prevention. https://www.cdc.gov/hiv/risk/art/index.html accessed: 22.01.2019 Coetzee, J., Jewkes, R., & Gray, G. E. (2017). Cross-sectional study of female sex workers in Soweto, South Africa : Factors associated with HIV infection. PLOS One, 1–21. https://doi.org/https://doi. org/10.1371/journal.pone.0184775 Cross, D., Barnes, A., Papageorgiou, A., Hadwen, K., Hearn, L., & Lester, L. (2015). A social–ecological framework for understanding and reducing cyberbullying behaviours. Aggression and Violent Behaviour, 23, 109-117. http://dx.doi.org/10.1016/j.avb.2015.05.016 Culture, S., August, N., & Taylor, P. (2010). “As prostitutes , we control our bodies”: perceptions of health and body in the lives of establishment-based female sex workers in 66 University of Ghana http://ugspace.ug.edu.gh Tijuana , Mexico Author ( s ): Shonali M . Choudhury Stable URL : https://www.jstor.org/stable/25681064 Mexico ’ As prosti. Culture, Health and Sexuality, 12(6), 677–689. https://doi.org/10.1080/13691051003797263 Decker, M. R., Crago, A.-L., Chu, S. K. H., Sherman, S. G., Seshu, M. S., Buthelezi, K., … Beyrer, C. (2015). Human rights violations against sex workers: burden and eff ect on HIV. The Lancet, 385. https://doi.org/10.1016/S0140-6736(14)60800-X Decker, M. R., Lyons, C., Billong, S. C., Njindam, I. M., Grosso, A., Nunez, G. T., … Baral, S. (2015). Gender-based violence against female sex workers in Cameroon : prevalence and associations with sexual HIV risk and access to health services and justice, (May). https://doi.org/10.1136/sextrans-2015-052463 Deering, K. N., Amin, A., Shoveller, J., Nesbitt, A., Gracia-Moreno, C., Duff, P., … Shannon, K. (2014). A Systematic Review of the Correlates of Violence Against Sex Workers. American Journal of Public Health, 104(5), 42–54. https://doi.org/10.2105/AJPH.2014.301909 Deering, K. N., Bhattacharjee, P., Bradley, J., Moses, S. S., Shannon, K., Shaw, S. Y., … Alary, M. (2011). Condom use within non-commercial partnerships of female sex workers in southern India. BMC Public Health, 11(Suppl 6). https://doi.org/10.1186/1471-2458-11-S6-S11 Dunkle, K. L., & Decker, M. R. (2013). Gender-Based Violence and HIV : Reviewing the Evidence for Links and Causal Pathways in the General Population and High-risk Groups. American Journal of Reproductive Immunology, 69, 20–26. https://doi.org/10.1111/aji.12039 Dunkle, K. L., Jewkes, R. K., Brown, H. C., Gray, G. E., Mcintryre, J. A., & Harlow, S. D. (2017). Gender-based violence , relationship power , and risk of HIV infection in women attending antenatal clinics in South Africa. Journal of Interpersonal Violence, 363. https://doi.org/10.1177/0886260517690875 Fawole, O. I., & Dagunduro, A. T. (2014). Prevalence and correlates of violence against female sex workers in Abuja , Nigeria. African Health Sciences, 14(2), 299–313. https://doi.org/http://dx.doi.org/10.4314/ahs.v14i2.4 GAC (2013) National HIV and AIDS, STI Policy. Ghana AIDS Commission, Accra GAC (2015). Bio-Behavioral Survey among Female Sex Workers and their Non-Paying Partners in Ghana. GAC (2016). National HIV and AIDS Strategic Plan 2016 – 2020, Ghana AIDS Commission, Accra GAC (2018). National and Sub-National HIV and AIDS Estimates and Projections, 2017 (draft). Ghana AIDS Commission, Accra George, A., Sabarwal, S., Martin, P., George, A., Sabarwal, S., & Martin, P. (2011). Violence in Contract Work Among Female Sex Workers in Andhra Pradesh , India. The Journal of Infectious Diseases, 204(December 2011), s1235–s1240. Retrieved from https://www.jstor.org/stable/41329793 Ghana Statistical Service (2012) 2010 Population and Housing Census: Summary Report of Final Results Accra: Ghana Statistical Service Goldenberg, S. M., Chettiar, J., Simo, A., Silverman, J. G., Strathdee, S. A., Montaner, J., & Shannon, K. (2015). Early sex work initiation independently elevates odds of HIV infection and police arrest among adult sex workers in a Canadian setting. Journal of Acquired Immune Deficiency Syndromes, 65(1), 122–128. https://doi.org/10.1097/QAI.0b013e3182a98ee6 Heise, L. (2011). What works to prevent partner violence? An evidence overviews. London, England: Department for International Development. Hesketh, T., Zhang, J., & Qiang, D. J. (2005). HIV knowledge and risk behaviour of female sex workers in Yunnan Province, China: potential as bridging groups to the general population. AIDS Care, 17(8), 958–966. https://doi.org/10.1080/09540120500100676 Jin, X., Smith, K., Chen, R. Y., Ding, G., Yao, Y., Wang, H., … Wang, N. (2011). HIV prevalence and risk behaviors among male clients of female sex workers in Yunnan, China. Acquired Immune Deficiency Syndrome, 53(27), 131–135. 67 University of Ghana http://ugspace.ug.edu.gh https://doi.org/10.1097/QAI.0b013e3181b3552a Le, M.-N., D’Onofrio, C. N., & Rogers, J. D. (2010). HIV Risk Behaviors among Three Classes of Female Sex Workers in Vietnam. The Journal of Sex Research, 47(1), 38–48. https://doi.org/10.1080/00224490902999278 Lima, F. S. da S., Merchan-Hammann, E., Urdaneta, M., Damacena, G. N., & Szwarcwald, C. L. (2017). Factors associated with violence against female sex workers in ten Brazilian cities Fatores associados à violência contra mulheres profissionais do sexo de dez cidades brasileiras Factores asociados a la violencia contra mujeres profesionales del sexo de. Critical Social Policy, 33(2), 1–15. https://doi.org/10.1590/0102- 311X00157815 Lin, T., Chen, V. C. H., Lee, C., Chen, C., Ferri, C. P., Gossop, M., & Lee, C. T. C. (2013). Prevalence , correlates , and risk perception of HIV infection among heroin users in Central Taiwan. Kaohsiung Journal of Medical Sciences, 29(12), 673–679. https://doi.org/10.1016/j.kjms.2013.04.003 Lyons, C. E., Grosso, A., Drame, F. M., Ketende, S., Diouf, D., Ba, I., … Columbia, B. (2017). Physical and sexual violence affecting female sex workers in Abidjan, Côte d’Ivoire: Prevalence, and the relationship with the work environment, HIV and access to health services Carrie, 75(1), 9–17. https://doi.org/10.1097/QAI.0000000000001310.Physical Marsh, D., & Stoker, G. (2010), Theory and methods in political science (3rd Edition),Palgrave Macmillan, London and New York McCloskey, A. L., Williams, C., Larsen, U., & Ann, B. L. (2005). Gender Inequality and Intimate Partner Violence among Women in Moshi , Tanzania. International Family Planning Perspective, 31(3), 124–130. Retrieved from https://www.jstor.org/stable/3649516 Mchugh, M. L. (2013). Lessons in biostatistics The Chi-square test of independence. Biochemia Medica, 23(2), 143–149. https://doi.org/dx.doi.org/10.11613/BM.2013.018 Mooney, A., Kidanu, A., Bradley, H. M., Kumoji, E. K., Kennedy, C. E., & Kerrigan, D. (2013). Work-related violence and inconsistent condom use with non-paying partners among female sex workers in Adama City , Ethiopia. BMC Public Health, 13(171), 1– 10. https://doi.org/10.1186/1471-2458-13-771 Ohshige, K., Morio, S., Mizushima, S., Kitamura, K., Tajima, K., Ito, A., … Soda, K. (2000). Cross-Sectional Study on Risk Factors of HIV among Female Commercial Sex Workers in Cambodia. Epidemiology and Infection, 124(1), 143–152. Retrieved from https://www.jstor.org/stable/3865030 Onyango, M. A., Adu-sarkodie, Y., Agyarko-poku, T., Asafo, M. K., Sylvester, J., Wondergem, P., … Beard, J. (2015). “ It ’ s All About Making a Life ”: Poverty , HIV , Violence , and Other Vulnerabilities Faced by Young Female Sex Workers in Kumasi , Ghana. Journal of Acquired Immune Deficiency Syndromes, 68, S131–S137. https://doi.org/10.1097/QAI.0000000000000455 Patel, S. K., Ganju, D., Prabhakar, P., & Adhikary, R. (2016). Relationship between mobility , violence and major depression among female sex workers : a cross-sectional study in southern India. BMJ Open, 6(e011439), 1–8. https://doi.org/10.1136/bmjopen-2016- 011439 Patterson, T. L., Goldenberg, S., Gallardo, M., Lozada, R., Shirley, J., Orozovich, P., … Strathdee, S. A. (2009). Correlates of HIV, STIs and Associated High Risk Behaviors among Male Clients of Female Sex Workers in Tijuana, Mexico. THe Lancet. HIV, 23(13), 1765–1771. https://doi.org/10.1097/QAD.0b013e32832f08a1.Correlates Pillai, S., Seshu, M., & Shivdas, M. (2008). Embracing the Rights of People in Prostitution and Sex Workers , to Address HIV and AIDS Effectively Embracing the rights of people in prostitution and sex workers , to address HIV and AIDS effectively. Gender and Development, 16(2). https://doi.org/10.1080/13552070802120491 Platt, L., Grenfell, P., Bonell, C., Creighton, S., Wellings, K., Parry, J., & Rhodes, T. (2011a). Risk of sexually transmitted infections and violence among indoor-working female sex workers in London : the effect of migration from Eastern Europe Demographic 68 University of Ghana http://ugspace.ug.edu.gh characteristics. The BMJ, 377–384. https://doi.org/10.1136/sti.2011.049544 Platt, L., Grenfell, P., Bonell, C., Creighton, S., Wellings, K., Parry, J., & Rhodes, T. (2011b). Risk of sexually transmitted infections and violence among indoor-working female sex workers in London : the effect of migration from Eastern Europe Demographic characteristics. Sexually Transmitted Infections, 87, 377–384. https://doi.org/10.1136/sti.2011.049544 Platt, L., Jolley, E., Rhodes, T., Hope, V., Latypov, A., Reynolds, L., & Wilson, D. (2013). Factors mediating HIV risk among female sex workers in Europe : a systematic review and ecological analysis. BMJ Open, 3(e002836). https://doi.org/10.1136/bmjopen-2013- 002836 Reed, E., Jhumka, G., Biradavolu, M., Vasavi, D., & Kim, M. B. (2010). The Context of Economic Insecurity and Its Relation to Violence and Risk Factors for HIV Among Female Sex Workers in Andhra Pradesh , India. Public Health Reports, 125, 81–89. https://doi.org/10.1177/00333549101250S412 Rhodes, T., Simic, M., Baros, S., Platt, L., & Zikic, B. (2008). Police violence and sexual risk among female and transvestite sex workers in Serbia : qualitative study. BMJ, 337(a811), 1–6. https://doi.org/10.1136/bmj.a811 Santo Gomes do Espirito, M., & Etheredge, G. D. (2005). Male clients of brothel prostitutes as a bridge for HIV infection between high risk and low risk groups of women in Senegal. STI, 342–344. https://doi.org/10.1136/sti.2004.011940 Sarkar, K., Bal, B., Mukherjee, R., Niyogi, S. K., Saha, M. K., & Bhattacharya, S. K. (2005). Epidemiology of HIV Infection among Brothel-based Sex Workers in Kolkata , India. Journal of Health, Population and Nutrition, 23(3), 231–235. Retrieved from https://www.jstor.org/stable/23499323 Schwitters, A., Swaminathan, M., Serwadda, D., Muyonga, M., Shiraishi, R. W., Benech, I., … Hladik, W. (2015). Prevalence of Rape and Client-Initiated Gender-Based Violence Among Female Sex Workers : Kampala , Uganda , 2012. AIDS Behav, 68–76. https://doi.org/10.1007/s10461-014-0957-y Semple, S. J., Stockman, J. K., Pitpitan, E. V, & Strathdee, S. A. (2015). Prevalence and Correlates of Client- Perpetrated Violence against Female Sex Workers in 13 Mexican Cities. PLOS One. https://doi.org/10.1371/journal.pone.0143317 Shannon, K, Strathdee, S., Goldenberg, S., Duff, P., Mwagi, P., Rusakova, M., … Boily, M.- C. (2016). Global Epidemiology of HIV among Female Sex Workers: Influence of Structural Determinants. Lancet (London, England), 385(9962), 55–71. https://doi.org/10.1016/S0140-6736(14)60931-4 Shannon, Kate, Goldenberg, S. M., Deering, K. N., & Strathdee, S. A. (2015). HIV Infection among Female Sex Workers in Concentrated and High Prevalence Epidemics: Why a Structural Determinant Framework is Needed. PubMed Central, 9(2), 174–182. https://doi.org/10.1097/COH.0000000000000042 Shannon, Kate, Kerr, T., Strathdee, S. A., Montaner, J. S., & Tyndall, M. W. (2009). Prevalence and structural correlates of gender based violence among a prospective cohort of female sex workers. BMJ, (339). https://doi.org/10.1136/bmj.b2939 Sherwood, J. A., Grosso, A., Decker, M. R., Peitzmeier, S., Papworth, E., Diouf, D., … Baral, S. (2015). Sexual violence against female sex workers in The Gambia : a cross-sectional examination of the associations between victimization and reproductive , sexual and mental health. Biomed Central, 1–10. https://doi.org/10.1186/s12889-015-1583-y Silverman, J. G., Raj, A., Cheng, D. M., Decker, M. R., Bridden, C., Pardeshi, M., … Samet, J. H. (2011). Sex Trafficking and Initiation-Related Violence , Alcohol Use, and HIV Risk Among HIV-infected Female Sex Workers in Mumbai, India. The Journal of Infectious Diseases, 204(December 2011), S1229–S1234. https://doi.org/10.1093/infdis/jir5409 Strathdee, S. A., West, B. S., Reed, E., Moazan, B., Azim, T., & Dolan, K. (2015). Substance Use and HIV Among Female Sex Workers and Female Prisoners: Risk Environments and Implications for Prevention, Treatment, and Policies. Journal of Acquired Immune Deficiency Syndromes, 69(0 1). 69 University of Ghana http://ugspace.ug.edu.gh https://doi.org/10.1097/QAI.0000000000000624.Substance Swain, S. N., Saggurti, N., Battala, M., Verma, R. K., & Jain, A. K. (2011). Experience of violence and adverse reproductive health outcomes , HIV risks among mobile female sex workers in India. BMC Public Health, 11(1), 357. https://doi.org/10.1186/1471- 2458-11-357 Terry, M. S. (2014). Women ’ s Health Care Applying the Social Ecological Model to Violence against Women with Disabilities, 3(6). https://doi.org/10.4172/2167- 0420.1000193 Toller, J., Reed, E., Blankenship, K. M., The, S., Diseases, I., Sex, S., … Blankenship, K. M. (2011). Police-Related Experiences and HIV Risk Among Female Sex Workers in Andhra Pradesh , India. The Journal of Infectious Diseases, 204(December 2011), S1223–S1228. https://doi.org/10.1093/infdis/jir539 Tounkara, F. K., Diabaté, S., Guédou, F. A., Ahoussinou, C., Kintin, F., Zannou, D. M., … Alary, M. (2014). Violence , condom breakage and HIV infection among female sex workers in Benin , West Africa. Sexually Transmitted Diseases, 41(5), 312–318. https://doi.org/10.1097/OLQ.0000000000000114 Ulibarri, M. D., Strathdee, S. A., Lozada, R., Staines-Orozco, H. S., Abramovitz, D., Semple, S., … Petterson, T. L. (2013). Condom use among female sex workers and their non- commercial partners: Effects of a sexual risk intervention in two Mexican cities. International Journal of STD $ AIDS, 23(4), 229–234. https://doi.org/10.1258/ijsa.2011.011184 USAID, (2017) Key Populations: Targeted Approaches Toward An Aids-Free Generation https://www.usaid.gov/what-we-do/global-health/hiv-and-aids/technical-areas/key- populations UNAIDS. (2018). 2017 Global HIV Statistics, (July), 5. Retrieved from http://www.unaids.org/sites/default/files/media_asset/UNAIDS_FactSheet_en.pdf World Health Organisation, (2012) Prevention and Treatment of HIV and other Sexually Transmitted Infections for Sex Workers in Low- and Middle-income Countries http://apps.who.int/iris/bitstream/handle/10665/77745/9789241504744_eng.pdf;jsession id=BDA2AEAF5CA4730765200A61F2131734?sequence=1 accessed: 16.01.2019 World Health Organisation (2018). HIV/AIDS – Key Facts https://www.who.int/news- room/fact-sheets/detail/hiv-aids accessed: 18.01.2019 Wirtz, A. L., Schwartz, S., Ketende, S., Anato, S., Nadedjo, F. D., Ouedraogo, H. G., … Baral, S. (2015). Sexual Violence , Condom Negotiation , and Condom Use in the Context of Sex Work : Results From Two West African Countries. Journal of Acquired Immune Deficiency Syndromes, 68, 171–179. Yodanis, C. L. (2004). Gender Inequality , Violence Against Women , and Fear: A Cross- National Test of the Feminist Theory of Violence Against Women. Journal of Interpersonal Violence, 19(6), 655–675. https://doi.org/10.1177/0886260504263868 70