Incarceration and exposure to internally displaced persons camps associated with reproductive rights abuses among sex workers in northern Uganda Margaret Erickson,1,2 Shira M Goldenberg,1,3 Monica Akello,4 Godfrey Muzaaya,4 Paul Nguyen,1 Josephine Birungi,5 Kate Shannon1,6 ABSTRACT Background While female sex workers (FSWs) face a high burden of violence and criminalisation, coupled with low access to safe, non-coercive care, little is known about such experiences among FSWs in conflict-affected settings, particularly as they relate to sexual and reproductive health (SRH) and rights. We explored factors associated with lifetime abortions among FSWs in northern Uganda; and separately modelled the independent effect of lifetime exposures to incarceration and living in internally displaced persons (IDP) camps on coerced and unsafe abortions. Methods Analyses are based on a community- based cross-sectional research project in Gulu District, northern Uganda (2011–2012) with The AIDS Support Organization (TASO) Gulu, FSWs, and other community organisations. We conducted questionnaires, sex worker/ community-led outreach to sex work venues, and voluntary HIV testing by TASO. Results Of 400 FSWs, 62 had ever accessed an abortion. In a multivariable model, gendered violence, both childhood mistreatment/or abuse at home [adjusted odds ratio (AOR) 1.96; 95% confidence interval (95% CI) 0.99–3.90] and workplace violence by clients (AOR 3.57; 95% CI 1.31–9.72) were linked to increased experiences of abortion. Lifetime exposure to incarceration retained an independent effect on increased odds of coerced abortion (AOR 5.16; 95% CI 1.39–19.11), and living in IDP camps was positively associated with unsafe abortion (AOR 4.71; 95% CI 1.42–15.61). Discussion and conclusions These results suggest a critical need for removal of legal and social barriers to realising the SRH rights of all women, and ensuring safe, voluntary access to reproductive choice for marginalised and criminalised populations of FSWs. BACKGROUND The World Health Organization (WHO) estimates that of an estimated 22 million unsafe abortions performed annually, 98% occur in low- and middle-income countries.1 In Uganda, an estimated one in five pregnancies end in an induced abortion.2 Abortion is only legal in Uganda under certain circumstances, resulting in the practice of unsafe and potentially dangerous methods of abor- tion for women.2 3 Unsafe abortion as Key message points ▸ Experience of abuse in childhood and in the workplace are associated with increased abortions among female sex workers (FSWs) in northern Uganda. ▸ FSWs frequently experience denied or impeded access to sexual and repro- ductive healthcare, representing a major infringement on their reproduct- ive rights. ▸ Living in internally displaced persons camps was independently associated with an increased risk of unsafe abor- tion, while experience of being incar- cerated was associated with coerced abortion.      201Erickson M, et al. J Fam Plann Reprod Health Care 2017;43:201–209. doi:10.1136/jfprhc-2016-101492 ►► Additional material is published online only. To view please visit the journal online (http://​dx.​doi.​org/​10.​1136/​ jfprhc-​2016-​101492). For numbered affiliations see end of article. Correspondence to Dr Kate Shannon, Department of Medicine, University of British Columbia, Gender and Sexual Health Initiative, B.C. Centre for Excellence in HIV/AIDS, St Paul’s Hospital, 608–1081 Burrard Street, Vancouver, British Columbia, Canada V6Z 1Y6; ​gshi@​cfenet.​ubc.​ca Received 1 March 2016 Revised 18 November 2016 Accepted 16 January 2017 Published Online First 9 February 2017 To cite: Erickson M, Goldenberg SM, Akello M, et al. J Fam Plann Reprod Health Care 2017;43:201–209. Research copyright. on D ecem ber 13, 2021 by guest. P rotected by http://jfprhc.bm j.com / J F am P lann R eprod H ealth C are: first published as 10.1136/jfprhc-2016-101492 on 9 F ebruary 2017. D ow nloaded from http://jfprhc.bmj.com/ http://www.fsrh.org/home/ http://crossmark.crossref.org/dialog/?doi=10.1136/jfprhc-2016-101492&domain=pdf&date_stamp=2017-4-12 http://jfprhc.bmj.com/ defined by the WHO is a “procedure for terminating an unwanted pregnancy either by persons lacking the necessary skills or in an environment lacking minimal medical standards or both”.1 Unsafe abortion remains the leading cause of maternal injuries and death within Uganda,2 with the driving factor for these alarmingly high rates being lack of safe abortion services.1 Marginalised populations such as female sex workers (FSWs) and women exposed to armed con- flict may be at increased risk for negative sexual and reproductive health (SRH) outcomes.4–6 In sub-Saharan Africa, women in sex work are already disproportionately affected by health and social inequalities such as a high burden of HIV,7–9 sexually transmitted infections (STIs),10 physical and sexual violence,7 and limited access to SRH services such as contraception.11 12 FSWs in sub-Saharan Africa have a pooled odds of HIV infection that is 12-fold higher than among the general female population.11 Importantly, structural determinants such as the crim- inalisation of sex work, stigma and discrimination, high levels of violence and human rights abuses, and conflict experiences have been shown to place FSWs at increased risk for negative health outcomes such as HIV/STIs.8 13 14 Of particular concern, recent evi- dence has found denial of HIV services for FSWs, as well as forced and mandatory HIV and STI testing, raising strong concern of coercive HIV care for this criminalised population.15 16 While most prior research has examined access to HIV/STI prevention, and to a lesser extent HIV care among FSWs, there remains a gap in research when it comes to broader SRH services for FSWs, including safe and voluntary access to abortions. Northern Uganda recently emerged from a two-decade-long conflict with the government’s Uganda’s People Defense Forces and the rebel Lord’s Resistance Army (LRA). Gross human rights violations including abduction of children and youth by the LRA have devastated the population, resulting in a break- down in social structures, forced migration, and large-scale displacement.17 Previous research suggests that exposure to conflict may disrupt or undermine access to SRH services (e.g. HIV testing, antiretroviral therapy).18 Prior studies have shown that FSWs living within post-conflict settings experience large-scale physical and sexual violence as a result of exposure to conflict, abduction by rebel forces, and experiences living in internally displaced persons (IDP) camps.19 However, there remains a gap in evidence regarding the links between conflict exposure and SRH and rights among FSWs, including FSWs’ uptake and experiences with abortion services. Additionally, the criminalisation of sex work in sub-Saharan Africa and other settings has been shown to result in human rights violations which contribute to negative SRH outcomes for FSWs.10 20 The limited research available on women’s experiences within the criminal justice system in sub-Saharan Africa demon- strates how current punitive approaches may under- mine SRH and rights for incarcerated female inmates.21 While there is a lack of literature specific to FSWs’ experiences within the criminal justice system within sub-Saharan Africa, research on criminalisation in other contexts (e.g. exposure to police harassment and arrest) indicates that the criminalisation of sex work is linked to serious rights violations, including denial of health services, as well as increased vulner- ability to negative SRH outcomes among FSWs.16 For example, research from other low- and middle-income countries has shown that incarcerated FSWs are highly marginalised and experience disproportionate negative SRH outcomes.22 Despite the high burden of violence, other human rights violations and health inequities faced by FSWs in sub-Saharan Africa, there remains a lack of research regarding FSWs’ SRH and rights, including access to safe and non-coercive abortion services, particularly within conflict-affected settings. The objective of this study was to examine structural contexts associated with lifetime abortions amongst FSWs in post-conflict Gulu, northern Uganda; and examine whether there is an independent confounding effect of historical exposure to incarceration and living in an IDP camp on coerced and unsafe abortions. METHODS This analysis drew on data from the Gulu Sexual Health Project, a community-based cross-sectional study in Gulu District, northern Uganda (May 2011– January 2012). The study was conducted by Canadian and Ugandan researchers in partnership with The AIDS Support Organization (TASO) and other local community partners. Service providers, FSWs and policy experts provided extensive input and guidance on the project and questionnaire. The study received ethical approval from the University of British Columbia Behavioural Research Ethics Board, TASO Institutional Review Board and is registered at the Ugandan National Council for Science and Technology. As described previously,10 a total of 400 FSWs were recruited through sex worker-led outreach to bars and other sex work venues (lodges, hotels and truck stops), as well as TASO-led outreach to former IDP camps, and referral from local community agencies (e. g. Gulu Refugee Committee). For outreach to bars and other sex work venues, sampling was based on ethnographic mapping and outreach planning by sex worker/peer outreach and TASO team. Time-location sampling was used as a method to enrol this hard to reach population, with physical spaces rather than persons as the primary sampling unit. Times and places where FSWs congregated were decided on based extensive outreach and engagement with FSWs Erickson M, et al. J Fam Plann Reprod Health Care 2017;43:201–209. doi:10.1136/jfprhc-2016-101492202 Research copyright. on D ecem ber 13, 2021 by guest. P rotected by http://jfprhc.bm j.com / J F am P lann R eprod H ealth C are: first published as 10.1136/jfprhc-2016-101492 on 9 F ebruary 2017. D ow nloaded from http://jfprhc.bmj.com/ and community agencies to better understand local sex work hot-spots. These methods are consistent with recruitment of other cohorts of FSWs and hidden/mobile populations.23 24 Eligibility criteria included: ≥14 years old, exchanged sex for money or resources (e.g. food, cell phone air time, clothing, shelter, etc.) in the previous 30 days, and able to provide informed consent. The participant response rate was >95%. Interviewer-administered questionnaires were con- ducted by trained female Acholi research assistants at the TASO-Gulu Clinic, or at the confidential location of the participant’s choosing (e.g. home, workplace). The questionnaire covered topics including sociode- mographics, sex work working conditions, trauma and violence, stigma, and sexual, reproductive, mental, and physical health and wellbeing, including access to HIV testing and treatment. Study participants received a standard monetary reimbursement of 10 000 Ugandan Shillings (UGX) (3.7 Canadian Dollars). Trained research assistants offered voluntary HIV testing and counselling, although this was not required for study participation. Referrals for food security pro- grammes and other health services were provided. Individual characteristics included age, place of birth and HIV serostatus. Variables related to repro- ductive sexual health and interpersonal factors included contraceptive use, history of STI and HIV tests, and types of partners (e.g. having an intimate non-commercial partner). Participants who had been pregnant were asked about abortion history, methods and decision-making. Unsafe abortion was determined by the type of abortion reported (i.e. surgical from a clinic, oral induction from a provider, vaginal method from a provider, traditional provider, or self- performed). Unsafe vs. safe abortion were determined by any type of abortion with the exclusion of ‘surgical from a clinic’. Although the authors acknowledge that ‘safe’ abortions can be provided orally from a pro- vider, no medically approved or ‘safe’ oral induction methods were reported by participants. Coerced abor- tion was defined as having an abortion not of your own choice or will (i.e. forced or convinced to undergo an abortion). Key structural variables used in this analysis included history of incarceration, recent physical/ sexual violence by clients, ever lived in an IDP camp, and ever abducted into the LRA. Specific variables examining historical violence (e.g. abuse/mistreatment in the home, history of physical and sexual assault, war-related exposure to violence, and experience with police/soldier harassment) were also examined. For our analysis of structural determinants of life- time abortion, the sample was restricted to women who had ever been pregnant (n=315). The modelling approach used in this article was two-fold. We first calculated descriptive statistics for individual, partner/interpersonal, and structural factors for women who had experienced one or more abortions. Wilcoxon rank-sum test for continuous vari- ables and Pearson’s Chi-squared (χ2) test (or Fisher’s exact test) were used for binary variables in the bivari- ate analysis. Variables which were hypothesised a priori to be related to abortions and which had p<0.10 in bivariate analyses were considered for inclu- sion in the explanatory multivariable logistic regres- sion model. Variables considered in the multivariate analysis included key individual (e.g. age, HIV status, education), SRH (e.g. condom use and access, STI testing) and structural (e.g. client-perpetrated violence, childhood abuse, lived in an IDP camp, incarceration) variables which were significant at p<0.10 in bivariate analysis. The final multivariate table is reflective of the variables that held in the model following the multi- variate analysis and which produced the model with the best fit. Model selection was performed using a backwards selection approach. Akaike’s information criterion (AIC) was used to determine the most parsi- monious model, as indicated by the lowest AIC value. Second, following the initial analysis for our main outcome of experience of abortion, sub-analyses were conducted exploring variables associated with two primary structural exposures: unsafe and coerced abortions. Separate multivariable confounder models were constructed using the approach of Maldonado and Greenland25 to examine the independent effect of historical exposure to (1) incarceration on coerced abortions and (2) living in IDP camps on unsafe abortions. While bivariate analyses examined the rela- tionship between exposure to IDP camps and incarcer- ation with both unsafe and coerced abortions, there was no significant association between exposure to IDP camps and coerced abortions, as well as incarcer- ation and unsafe abortion in the final analyses. Potential sources of bias were addressed by adjusting for potential confounders identified in bivariate ana- lysis. All potential confounders hypothesised a priori and with p<0.10 in bivariate analysis were included in multivariable analysis. Potential confounders were selected for inclusion in each respective multivariable model to examine the effect on the magnitude of the coefficient of the primary variable of interest. Starting with the full model of all available variables, reduced models were fit to compare the associated value of coefficient of the primary variable to the one in the full model. The secondary variable associated with the smallest relative change in the main independent vari- able was dropped. This iterative process continued until the minimum change exceeded 5%. Remaining variables were included as confounders in the final model. All analyses were performed using SAS V.9.3 (SAS, Cary, NC, USA). RESULTS Our analysis of abortions was restricted to 315 FSWs who had ever been pregnant at least once in their Erickson M, et al. J Fam Plann Reprod Health Care 2017;43:201–209. doi:10.1136/jfprhc-2016-101492 203 Research copyright. on D ecem ber 13, 2021 by guest. P rotected by http://jfprhc.bm j.com / J F am P lann R eprod H ealth C are: first published as 10.1136/jfprhc-2016-101492 on 9 F ebruary 2017. D ow nloaded from http://jfprhc.bmj.com/ lifetime. Of these women, 62 (19.7%) had experi- enced at least one abortion. Of 315 women who had ever been pregnant, participants had a median age of 23 [interquartile range (IQR) 20–26], with 38.4% (121) living with HIV and 292 (92.7%) reporting having at least one child. Some 208 (66.0%) women had little or no primary education (Table 1). Only 184 (58.4%) FSWs had ever used hormonal non-barrier contraceptives (birth control pills, Depo- Provera®/injectables, intrauterine devices, implants or sterilisation). Though the majority (n=282, 89.5%) of women reported lifetime use of condoms as contra- ceptive, 186 (59.1%) FSWs reported difficulty accessing condoms in the last 6 months. Only 126 (40.0%) FSWs had ever been taught about pregnancy growing up. In terms of structural exposures, slightly over one- quarter of participants (n=85, 27.0%) reported ever having been incarcerated. FSWs experienced high levels of violence by clients, with 249 (79.1%) FSWs reporting physical and/or sexual violence from clients in the last 6 months. Of 315 FSWs, 211 (67.0%) had lived in an IDP camp, and 109 (34.6%) had been abducted into the LRA. Wartime-related violence was high, with 179 (56.8%) women reporting receiving physical threats, 100 (31.8%) women reporting severe physical violence, and 67 (21.3%) women reporting war-related rape. General experience with rape (including non-war-related rape) was much higher, with 207 (65.7%) FSWs reporting having been raped throughout their lifetime. Of 315 women, 186 (59.1%) reported childhood mistreatment and abuse. Lifetime physical violence was reported by 262 (83.2%) FSWs. In a multivariable model (Table 2), women who reported ever using condoms for contraception [adjusted odds ratio (AOR) 0.36; 95% confidence interval (95% CI) 0.14–0.81] and who had children (AOR 0.30; 95% CI 0.10–0.85) were less likely to have undergone an abortion in their lifetime. Women who were more financially stable (AOR 2.00; 95% CI 1.05–3.80), experienced childhood mistreatment or abuse at home (AOR 1.96; 95% CI 0.99–3.90), and who experienced work-related physical/sexual vio- lence by clients (AOR 3.57; 95% CI 1.31–9.72) were more likely to have an abortion. Of 62 FSWs who had had at least one abortion, 64.5% (n=40) had undergone an unsafe abortion, of which the majority (75.0%) were self-performed (e.g. in the woman’s home using herbs or hormonal drugs). Additionally, almost one-third (29.0%, n=18) of women with a history of abortion reported having been coerced into having an abortion. In a separate adjusted multivariable confounder model (Table 3), historical exposure to living in an IDP camp retained an association with increased odds of unsafe abortion (AOR 4.71; 95% CI 1.42–15.61), while historical incarceration retained an independent association with coerced abortion (AOR 5.16; 95% CI 1.39– 19.11). DISCUSSION In this study, experience with both childhood abuse and workplace violence were strongly associated with increased likelihood of abortions among FSWs in the conflict-affected Gulu region of northern Uganda. Consistent with work from other countries as well as the WHO multi-country studies,26–28 our findings suggest that physical and sexual violence may be driving unintended or unwanted pregnancies among this population of women. Although access to SRH services including abortions are critical to reproduct- ive rights and choice for all women, legal and social barriers in Uganda, including the criminalised and stigmatised nature of both sex work and abor- tions,2 10 29 place serious constraints on reproductive rights and access to safe, non-coercive abortion ser- vices for marginalised women. Of particular concern, sub-analyses conducted as part of this study documented unacceptable high rates of unsafe and coerced abortions among FSWs in northern Uganda. Our findings revealed that exposure to living in an IDP camp remained independently associated with almost a five-fold increased risk for having an unsafe abortion. Unfortunately, as shown in previous research, access to safe and non-coercive abortion services has often been left out of reproduct- ive health care for refugee and internally displaced women.5 Additionally, discussions around gender- based violence are lacking when it comes to advocat- ing for human rights and access to SRH care within IDP camps in sub-Saharan Africa.30 Our findings echo the need to scale up access to reproductive health ser- vices within IDP camps, including access to safe and legal abortions. Of additional concern, almost one-third of abortions were coerced, suggesting major human rights infringe- ments on FSWs’ reproductive rights and choice, which has been documented as an increasing concern world- wide and particularly in sub-Saharan Africa.16 Exposure to incarceration was linked to a five-fold increased odds of having a coerced abortion, adding to evidence that criminalisation of sex work may lead to detrimental impacts on FSWs’ rights and health.14 Coerced abortion is often linked to inequalities in gender and power,31 and is a human rights violation on women’s bodies and reproductive choice, as is any coercive form of medical care or support. Our findings raise concern regarding the fact that the most margina- lised FSWs who already face persecution by the law appear to be the most likely to report coercion and a lack of control over decisions regarding their own sexual health. While previous reports have documen- ted large-scale human rights violations for women within prisons of sub-Saharan Africa,32 research spe- cific to FSWs or coercion related to SRH decisions and Erickson M, et al. J Fam Plann Reprod Health Care 2017;43:201–209. doi:10.1136/jfprhc-2016-101492204 Research copyright. on D ecem ber 13, 2021 by guest. P rotected by http://jfprhc.bm j.com / J F am P lann R eprod H ealth C are: first published as 10.1136/jfprhc-2016-101492 on 9 F ebruary 2017. D ow nloaded from http://jfprhc.bmj.com/ Table 1 Characteristics and unadjusted odds ratios related to lifetime abortion use among 315 female sex workers in Gulu, northern Uganda Characteristic Yes [n (%)] (n=62) No [n (%)] (n=254) OR (95% CI) p Individual characteristics Age (years) (Med, IQR) 21 (19–25) 23 (20–26) 0.93 (0.87–1.00) 0.045 Born in Gulu Yes 41 (66.1) 165 (65.2) 1.04 (0.58–1.87) 0.892 No 21 (33.9) 88 (34.8) Is an orphan Yes 51 (82.3) 214 (84.6) 0.85 (0.41–1.76) 0.653 No 11 (17.7) 39 (15.4) HIV-positive Yes 17 (27.4) 104 (41.1) 0.56 (0.30–1.04) 0.065 No 43 (69.4) 148 (58.5) None or incomplete primary education Yes 28 (45.2) 180 (71.2) 0.33 (0.19–0.59) <0.001 No 34 (54.8) 73 (28.9) Have children* Yes 54 (87.1) 238 (94.1) 0.43 (0.17–1.05) 0.096 No 8 (12.9) 15 (5.9) Average weekly income from sex work, 1000s UGX (Med, IQR) 40 (20–80) 40 (30–70) 1.00 (1.00–1.01) 0.948 Financially support others Yes 56 (90.3) 242 (95.6) 0.42 (0.15–1.20) 0.115 No 6 (9.7) 11 (4.4) Good current financial situation compared to neighbours* Yes 22 (35.5) 60 (23.7) 1.77 (0.98–3.21) 0.058 No 40 (64.5) 193 (76.3) SRH and interpersonal factors Ever used non-barrier contraceptives* Yes 42 (67.7) 142 (56.1) 1.64 (0.91–2.95) 0.096 No 20 (32.3) 111 (43.9) Ever used male condoms for pregnancy prevention* Yes 50 (80.7) 232 (91.7) 0.38 (0.17–0.82) 0.011 No 12 (19.4) 21 (8.3) Difficulty accessing condoms† Yes 33 (53.3) 153 (60.5) 0.74 (0.43–1.30) 0.298 No 29 (46.8) 100 (39.5) Ever been tested for STIs* Yes 22 (35.5) 124 (49.0) 0.57 (0.32–1.02) 0.056 No 40 (64.5) 129 (51.0) Taught about pregnancy as a child Yes 20 (32.3) 106 (41.9) 0.67 (0.37–1.21) 0.182 No 39 (62.9) 138 (54.6) Have an intimate (non-commercial) partner Yes 60 (96.8) 239 (94.5) 1.76 (0.39–7.94) 0.747 No 2 (3.2) 14 (5.5) Structural factors Ever been incarcerated* Yes 17 (27.4) 68 (26.9) 1.03 (0.55–1.92) 0.931 No 45 (72.6) 185 (73.1) Physical or sexual violence by clients† Yes 57 (91.94) 192 (75.89) 3.62 (1.39–9.45) 0.005 No 5 (8.06) 61 (24.11) Continued Erickson M, et al. J Fam Plann Reprod Health Care 2017;43:201–209. doi:10.1136/jfprhc-2016-101492 205 Research copyright. on D ecem ber 13, 2021 by guest. P rotected by http://jfprhc.bm j.com / J F am P lann R eprod H ealth C are: first published as 10.1136/jfprhc-2016-101492 on 9 F ebruary 2017. D ow nloaded from http://jfprhc.bmj.com/ services within criminal justice settings within this context is relatively non-existent. Reports have shown that on average only 2.2% of all inmates in sub-Saharan Africa are female, with Uganda having 3.4% female inmates out of the total prison popula- tion.32 Due to the low number of female inmates com- bined with a lack of resources, women and young girls are often detained alongside male inmates and are largely supervised entirely by male staff.32 Within incarceration settings where the majority of the detained population are male, SRH services for incar- cerated women are likely to be extremely limited, if at all available. Our findings shed light on a major human rights infringement on reproductive health for FSWs in sub-Saharan Africa surrounding the criminalisation of sex work and their subsequent incarceration. LIMITATIONS Although these findings provide a critical first step in highlighting the role of structural determinants for FSWs’ SRH and rights, several limitations and poten- tial biases should be considered in interpreting our findings. While it is not possible to examine temporal- ity in cross-sectional data, we do know that lifetime exposures of IDP camps and wartime abduction are historical in the context of the conflict, and as such may predate the unsafe abortions. Nevertheless, as a cross-sectional analysis, our findings cannot be used to infer causality; as it is possible that causal relation- ships could work in either direction, future longitu- dinal and intervention studies remain needed. Additionally, while we acknowledge that this study may not be representative of all FSWs, our team Table 1 Continued Characteristic Yes [n (%)] (n=62) No [n (%)] (n=254) OR (95% CI) p Fear of violence from client if asked to use a condom† Yes 48 (77.4) 187 (73.9) 1.37 (0.68–2.74) 0.373 No 12 (19.4) 64 (25.3) Used drugs or alcohol while working† Yes 46 (74.2) 161 (63.6) 1.64 (0.88–3.07) 0.117 No 16 (25.8) 92 (36.4) Ever lived in an IDP camp* Yes 43 (69.4) 168 (66.4) 1.15 (0.63–2.09) 0.658 No 19 (30.6) 85 (33.6) Ever been abducted into the LRA* Yes 14 (22.6) 95 (37.6) 0.49 (0.25–0.93) 0.026 No 48 (77.4) 158 (62.5) Historical violence Mistreated/abused at home* Yes 45 (72.6) 141 (55.7) 2.37 (1.24–4.55) 0.008 No 14 (22.6) 104 (41.1) Ever defiled/raped* Yes 37 (59.7) 170 (67.2) 0.67 (0.37–1.24) 0.201 No 21 (33.9) 65 (25.7) Ever experienced any police/soldier harassment* Yes 22 (35.5) 88 (34.8) 1.03 (0.58–1.84) 0.917 No 40 (64.5) 165 (65.2) Ever been verbally threatened* Yes 56 (90.3) 216 (85.4) 1.60 (0.64–3.98) 0.309 No 6 (9.7) 37 (14.6) Ever been sexually assaulted* Yes 46 (74.2) 159 (62.9) 1.70 (0.91–3.17) 0.093 No 16 (25.8) 94 (37.2) Ever been physically assaulted* Yes 52 (83.9) 210 (83.0) 1.07 (0.50–2.26) 0.870 No 10 (16.1) 43 (17.0) *Lifetime. †Last 6 months. CI, confidence interval; IDP, internally displaced persons; IQR, interquartile range; LRA, Lord’s Resistance Army; Med, median; OR, odds ratio; SRH, sexual and reproductive health; STIs, sexually transmitted infections; UGX, Ugandan Shilling. Erickson M, et al. J Fam Plann Reprod Health Care 2017;43:201–209. doi:10.1136/jfprhc-2016-101492206 Research copyright. on D ecem ber 13, 2021 by guest. P rotected by http://jfprhc.bm j.com / J F am P lann R eprod H ealth C are: first published as 10.1136/jfprhc-2016-101492 on 9 F ebruary 2017. D ow nloaded from http://jfprhc.bmj.com/ employed a variety of strategies to ensure that our sample was as representative as possible, including use of rigorous time-location sampling methods, strong community partnerships, and regular outreach efforts, which resulted in our ability to recruit a large and diverse sample of hidden and hard-to-reach FSWs across a diverse range of work environments and venues. Nonetheless, as time-location sampling has the potential to over-represent women who visit certain locations more often than another, it is pos- sible that our results may not be fully representative of all FSWs across all work environments, and may underestimate variance in some of the characteristics and associations reported. As the sample size for this study was based on preliminary estimates of HIV prevalence among young women in Uganda, and not based on experience with abortion, this may have limited our power to detect differences in experience with abortions. In light of the small number of cases used in the sub-analyses, the results and confidence intervals associated with our confounder models should be interpreted with caution. However, the small number of cases of coerced and unsafe abortions documented in this study would have made it more difficult to detect the associations we noted between incarceration, internal displacement, and these out- comes, biasing our findings towards the null. Building on findings of this exploratory analysis, more robust studies investigating patterns and determinants of coercive and unsafe abortions for FSWs remain critic- ally needed. CONCLUSIONS In sub-Saharan Africa, women in sex work face mul- tiple barriers when it comes to access to SRH services, representing a major infringement on their reproduct- ive rights. In this study, experience with historical and ongoing violence was linked to increased odds of abortions among FSWs in conflict-affected Gulu, northern Uganda. With almost a five-fold increased risk of unsafe abortions among those who lived in IDP camps, findings of sub-analyses show that there continues to be large gaps in care concerning access to safe reproductive services for refugee and displaced women in sex work. Furthermore, one-third of abor- tions were coerced – a major human rights infringe- ment on reproductive choice for FSWs. While access to non-coercive and safe abortion practices represents a human right fundamental to promoting women’s overall wellbeing, legal prohibitions of access to abor- tion services coupled with the criminalised and stig- matised nature of sex work in Uganda, as in other settings in sub-Saharan Africa, continue to pose major gaps in access to SRH care for FSWs. Given examples in other settings of denial of access to care among FSWs in sub-Saharan Africa and forced/coercive HIV testing,15 our findings echo broader human rights concerns of the harms of criminalisation and incarcer- ation on reproductive health care for FSWs. In order to protect and promote the health and rights of FSWs in northern Uganda as well as other conflict-affected settings in sub-Saharan Africa, strengthening rights-based SRH for marginalised and displaced women, including access to non-coercive and safe abortions services, remain critically needed. Future programmes and legal reforms remain critical in order to protect the SRH choice and rights for women, especially marginalised populations of FSWs. Author affiliations 1Gender and Sexual Health Initiative, British Columbia Centre for Excellence in HIV/AIDS, Vancouver, British Columbia, Canada Table 2 Multivariable logistic regression analysis of factors independently associated with lifetime abortion among female sex workers who had ever been pregnant (n=315) in Gulu, northern Uganda Variable AOR (95% CI) p Ever used male condoms* (Yes vs No) 0.34 (0.14–0.81) 0.015 Have children* (Yes vs No) 0.30 (0.10–0.85) 0.023 Mistreated/abused at home* (Yes vs No) 1.96 (0.99–3.90) 0.054 Good financial situation (Yes vs No) 2.00 (1.05–3.80) 0.036 Physical or sexual violence by clients‡ (Yes vs No) 3.57 (1.31–9.72) 0.013 *Lifetime. ‡Last 6 months. AOR, adjusted odds ratio; CI, confidence interval. Table 3 Separate confounder models examining the independent effects of lifetime incarceration and living in an internally displaced persons camp on coerced and unsafe abortion among female sex workers (n=62) in Gulu, northern Uganda* Outcome [AOR (95% CI)] Exposure Coerced abortion Unsafe abortion Lifetime incarceration† 5.16 (1.39–19.11) – Ever lived in an IDP camp‡ – 4.71 (1.42–15.61) *Bivariate analyses were conducted exploring variables associated with unsafe and coerced abortions in our sample (see online Supplementary Table A and B). Based on findings of bivariate analyses, multivariable confounder models were then constructed using the approach of Maldonado and Greenland25 to examine the independent effect of historical exposure to (1) incarceration on coerced abortions and (2) living in IDP camps on unsafe abortions. †Model selection included age and fear of violence when asking clients to use condoms as potential confounders; only age was retained in the final model. ‡Model selection included sexually transmitted infection testing, and fear of violence when asking clients to use condoms as potential confounders; however, no confounding variables were retained in the final model. AOR, adjusted odds ratio; CI, confidence interval; IDP, internally displaced persons. Erickson M, et al. J Fam Plann Reprod Health Care 2017;43:201–209. doi:10.1136/jfprhc-2016-101492 207 Research copyright. on D ecem ber 13, 2021 by guest. P rotected by http://jfprhc.bm j.com / J F am P lann R eprod H ealth C are: first published as 10.1136/jfprhc-2016-101492 on 9 F ebruary 2017. D ow nloaded from http://dx.doi.org/10.1136/jfprhc-2016-101492 http://dx.doi.org/10.1136/jfprhc-2016-101492 http://jfprhc.bmj.com/ 2School of Population and Public Health, University of British Columbia, Vancouver, British Columbia, Canada 3Faculty of Health Sciences, Simon Fraser University, Burnaby, British Columbia, Canada 4The AIDS Support Organization Gulu, Gulu, Uganda 5The AIDS Support Organization, Kampala, Uganda 6Department of Medicine, University of British Columbia, Vancouver, British Columbia, Canada Acknowledgements The authors wish to thank all those who contributed to this project, particularly participants, Acholi interview and peer/SWoutreach teams, TASO Gulu community advisory board and partner agencies. The authors acknowledge in particular G.M., Esther Achan, J.B., M.A., Miriam Ajok, Katie Muldoon, Jacky Ayat, Beatrice Baraka, Bibian Achan, Grace Lakot, Winnie Amunu, Andrew Mijumbi Ojok, Melissa Braschel and Sabina Dobrer for their research and administrative support. This research was supported through a CIHR Team Grant in Gender, Violence and Health. SG is supported by the Canadian Institutes of Health Research (CIHR) and the National Institute of Health. KS is supported by a Canada Research Chair in Global Sexual Health and HIV/ AIDS and the Michael Smith Foundation for Health Research. Funding Canadian Institutes of Health Research. This research was supported through a CIHR Team Grant in Gender, Violence and Health (TVG-115616). Competing interests None declared. Ethics approval The study received ethical approval from the University of British Columbia Behavioural Research Ethics Board, TASO Institutional Review Board and is registered at the Ugandan National Council for Science and Technology. Provenance and peer review Not commissioned; externally peer reviewed. http://jfprhc.bmj.com/site/about/licence.pdf REFERENCES 1 World Health Organization. Safe abortion: technical and policy guidance for health systems. Reprod Health Matters 2012;20:205–207. 2 Centre for Reproductive Rights, the International Women’s Human Rights Clinic at Georgetown Law and the O’Neill Institute for National and Global Health at Georgetown Law. The Stakes are High: The Tragic Impact of Unsafe Abortion and Inadequate Access to Contraception in Uganda. 2013. http:// tbinternet.ohchr.org/Treaties/CESCR/Shared%20Documents/ UGA/INT_CESCR_CSS_UGA_20279_E.pdf. 3 Polis CB, Lutalo T, Wawer M, et al. Coerced sexual debut and lifetime abortion attempts among women in Rakai, Uganda. Int J Gynaecol Obstet 2009;104:105–109. 4 Patel SH, Muyinda H, Sewankambo NK, et al. In the face of war: examining sexual vulnerabilities of Acholi adolescent girls living in displacement camps in conflict-affected Northern Uganda. BMC Int Health Hum Rights 2012;12:38. 5 Lehmann A. Safe abortion: a right for refugees? Reprod Health Matters 2002;10:151–155. 6 Mulumba D. Encampment of communities in war-affected areas and its effect on their livelihood security and reproductive health: the case of Northern Uganda. East Afr Soc Sci Res Rev 2011;27:107–129. 7 Fonner VA, Kerrigan D, Mnisi Z, et al. Social cohesion, social participation, and HIV related risk among female sex workers in Swaziland. PLoS ONE 2014;9:e87527. 8 Scorgie F, Chersich MF, Ntaganira I, et al. Socio-demographic characteristics and behavioral risk factors of female sex workers in sub-Saharan Africa: a systematic review. AIDS Behav 2012;16:920–933. 9 Peitzmeier S, Mason K, Ceesay N, et al. A cross-sectional evaluation of the prevalence and associations of HIV among female sex workers in the Gambia. Int J STD AIDS 2014;25:244–252. 10 Erickson M, Goldenberg SM, Ajok M, et al. Structural determinants of dual contraceptive use among female sex workers in Gulu, northern Uganda. Int J Gynaecol Obstet 2015;131:91–95. 11 Baral S, Beyrer C, Muessig K, et al. Burden of HIV among female sex workers in low-income and middle-income countries: a systematic review and meta-analysis. Lancet Infect Dis 2012;12:538–549. 12 Duff P, Shoveller J, Zhang R, et al. High lifetime pregnancy and low contraceptive usage among sex workers who use drugs – an unmet reproductive health need. BMC Pregnancy Childbirth 2011;11:61. 13 Bukenya J, Vandepitte J, Kwikiriza M, et al. Condom use among female sex workers in Uganda. AIDS Care 2013;25:767–774. 14 Shannon K, Strathdee SA, Goldenberg SM, et al. Global epidemiology of HIV among female sex workers: influence of structural determinants. Lancet 2015;385:55–71. 15 Scorgie F, Nakato D, Harper E, et al. “We are despised in the hospitals”: sex workers’ experiences of accessing health care in four African countries. Cult Health Sex 2013;15:450–465. 16 Decker MR, Crago A-L, Chu SKH, et al. Human rights violations against sex workers: burden and effect on HIV. Lancet 2015;385:186–199. 17 Human Rights Watch. Abducted and Abused: Renewed Conflict in Northern Uganda. Human Rights Watch Vol. 15, No. 12(A), 2003. 18 Mills EJ, Singh S, Nelson BD, et al. The impact of conflict on HIV/AIDS in sub-Saharan Africa. Int J STD AIDS 2006;17:713–717. 19 Muldoon KA, Muzaaya G, Betancourt TS, et al. After abduction: exploring access to reintegration programs and mental health status among young female abductees in Northern Uganda. Confl Health 2014;8:5. 20 Nyanzi S. Homosexuality, sex work, and HIV/AIDS in displacement and post-conflict settings: the case of refugees in Uganda. Int Peacekeeping 2013;20:450–468. 21 Johnson JE, Carney T, Kline T, et al. Incarceration history relative to health, substance use, and violence in a sample of vulnerable South African women: implications for health services in criminal justice settings. Subst Abuse Rehabil 2012;3 (Suppl. 1):59–69. 22 Tucker JD, Ren X. Sex worker incarceration in the People’s Republic of China. Sex Transm Infect 2008;84:34–35. 23 Goldenberg SM, Chettiar J, Simo A, et al. Early sex work initiation independently elevates odds of HIV infection and police arrest among adult sex workers in a Canadian setting. J Acquir Immune Defic Syndr 2014;65:122–128. 24 Clark JL, Konda KA, Silva-Santisteban A, et al. Sampling methodologies for epidemiologic surveillance of men who have sex with men and transgender women in Latin America: an empiric comparison of convenience sampling, time space sampling, and respondent driven sampling. AIDS Behav 2014;18:2338–2348. 25 Maldonado G, Greenland S. Simulation study of confounder- selection strategies. Am J Epidemiol 1993;138:923–936. Erickson M, et al. J Fam Plann Reprod Health Care 2017;43:201–209. doi:10.1136/jfprhc-2016-101492208 Research copyright. on D ecem ber 13, 2021 by guest. P rotected by http://jfprhc.bm j.com / J F am P lann R eprod H ealth C are: first published as 10.1136/jfprhc-2016-101492 on 9 F ebruary 2017. D ow nloaded from http://jfprhc.bmj.com/site/about/licence.pdf http://jfprhc.bmj.com/site/about/licence.pdf http://dx.doi.org/10.1016/S0968-8080(12)39623-7 http://dx.doi.org/10.1016/j.ijgo.2008.10.002 http://dx.doi.org/10.1016/j.ijgo.2008.10.002 http://dx.doi.org/10.1186/1472-698X-12-38 http://dx.doi.org/10.1016/S0968-8080(02)00026-5 http://dx.doi.org/10.1016/S0968-8080(02)00026-5 http://dx.doi.org/10.1353/eas.2011.0002 http://dx.doi.org/10.1353/eas.2011.0002 http://dx.doi.org/10.1371/journal.pone.0087527 http://dx.doi.org/10.1007/s10461-011-9985-z http://dx.doi.org/10.1177/0956462413498858 http://dx.doi.org/10.1016/j.ijgo.2015.04.029 http://dx.doi.org/10.1016/S1473-3099(12)70066-X http://dx.doi.org/10.1016/S1473-3099(12)70066-X http://dx.doi.org/10.1186/1471-2393-11-61 http://dx.doi.org/10.1186/1471-2393-11-61 http://dx.doi.org/10.1080/09540121.2012.748863 http://dx.doi.org/10.1016/S0140-6736(14)60931-4 http://dx.doi.org/10.1080/13691058.2012.763187 http://dx.doi.org/10.1016/S0140-6736(14)60800-X http://dx.doi.org/10.1186/1752-1505-8-5 http://dx.doi.org/10.1080/13533312.2013.846136 http://dx.doi.org/10.2147/SAR.S21351 http://dx.doi.org/10.1136/sti.2007.027235 http://dx.doi.org/10.1097/QAI.0b013e3182a98ee6 http://dx.doi.org/10.1007/s10461-013-0680-0 http://jfprhc.bmj.com/ 26 Russo NF, Denious JE. Violence in the lives of women having abortions: implications for practice and public policy. Prof Psychol Res Pract 2001;32:142–150. 27 Pallitto CC, García-Moreno C, Jansen HAFM, et al. Intimate partner violence, abortion, and unintended pregnancy: results from the WHO Multi-country Study on Women’s Health and Domestic Violence. Int J Gynaecol Obstet 2013;120:3–9. 28 Boden JM, Fergusson DM, Horwood LJ. Experience of sexual abuse in childhood and abortion in adolescence and early adulthood. Child Abuse Negl 2009;33:870–876. 29 Singh S, Moore A, Bankole A, et al. Unintended Pregnancy and Induced Abortion in Uganda: Causes and Consequences. 2006. https://www.guttmacher.org/sites/default/files/pdfs/pubs/2006/ 11/27/UgandaUPIA.pdf [accessed 4 November 2016]. 30 Whelan A, Blogg J. “Halfway people”: refugee views of reproductive health services. Glob Public Health 2007;2:373–394. 31 Braam T, Hessini L. The power dynamics perpetuating unsafe abortion in Africa: a feminist perspective. Afr J Reprod Health 2004;8:43–51. 32 Egamberdi N. HIV and Prisons in sub-Saharan Africa: Opportunities for Action. 2007. http://documents.worldbank. org/curated/en/2007/10/9110393/hiv-prisons-sub-saharan- africa-opportunities-action. Erickson M, et al. J Fam Plann Reprod Health Care 2017;43:201–209. doi:10.1136/jfprhc-2016-101492 209 Research copyright. on D ecem ber 13, 2021 by guest. P rotected by http://jfprhc.bm j.com / J F am P lann R eprod H ealth C are: first published as 10.1136/jfprhc-2016-101492 on 9 F ebruary 2017. D ow nloaded from http://dx.doi.org/10.1037/0735-7028.32.2.142 http://dx.doi.org/10.1037/0735-7028.32.2.142 http://dx.doi.org/10.1016/j.ijgo.2012.07.003 http://dx.doi.org/10.1016/j.chiabu.2009.04.006 http://dx.doi.org/10.1080/17441690701267048 http://dx.doi.org/10.2307/3583304 http://jfprhc.bmj.com/ Incarceration and exposure to internally displaced persons camps associated with reproductive rights abuses among sex workers in northern Uganda Abstract Background Methods Results Discussion Limitations Conclusions References