Content uploaded by Alexander Vu
Author content
All content in this area was uploaded by Alexander Vu on Mar 20, 2014
Content may be subject to copyright.
!
"!
The prevalence of sexual violence among female refugees in complex humanitarian
emergencies: a systematic review and meta-analysis.
Vu A, Adam A, Wirtz AL, Pham K, Rubenstein L, Glass N, Singh S
PLoS Currents Disasters. 2014 Mar 18. Edition 1.
doi: 10.1371/currents.dis.835f10778fd80ae031aac12d3b533ca7.
Corresponding Author Information
Alexander Vu, DO, MPH
Johns Hopkins University
International Emergency Medicine and Public Health Fellowship Program
School of Medicine, Department of Emergency Medicine
School of Public Health, Department of International Health
5801 Smith Avenue, Suite 3220, Davis Bldg, Baltimore, MD 21209
Office: 410-955-4059
avu3@jhmi.edu
Authors and affiliations:
Atif Adam, MBBS, PhD candidate
Department of International Health, Johns Hopkins Bloomberg School of Public Health, Baltimore, USA
aadam@jhsph.edu
Andrea Wirtz, MHS, PhD candidate
Department of Emergency Medicine, Johns Hopkins School of Medicine,
Center for Public Health and Human Rights, Department of Epidemiology, Johns Hopkins Bloomberg
School of Public Health, Baltimore, USA
awirtz@jhsph.edu
Kiemanh Pham, MD, MPH
International Emergency Medicine and Public Health Fellowship Program, Department of Emergency
Medicine, Johns Hopkins School of Medicine, Baltimore, USA
kpham4@jhmi.edu
Leonard Rubenstein, JD, LLM
Center for Public Health and Human Rights, Department of Epidemiology, Johns Hopkins Bloomberg
School of Public Health, Baltimore, USA
lrubenst@jhsph.edu
Nancy Glass, PhD, MPH, RN, FAAN
Johns Hopkins University School of Nursing, Health System, Department of International Health, Johns
Hopkins Bloomberg School of Public Health, Baltimore, USA
nglass1@jhu.edu
Sonal Singh, MD, MPH
Department of Medicine, Johns Hopkins School of Medicine
Center for Public Health and Human Rights, Department of Epidemiology, Johns Hopkins Bloomberg
School of Public Health, Baltimore, USA
sosingh@jhsph.edu
!
#!
Abstract:
Importance: Refugees and internally displaced populations (IDP) are highly vulnerable to sexual
violence during conflict and subsequent displacement. However, accurate estimates of the
prevalence of sexual violence among in these populations remain uncertain.
Objective: Our objective was to estimate the prevalence of sexual violence among refugees and
displaced populations in complex humanitarian emergencies.
Data Source: We conducted systematic review of relevant literature in multiple databases
(EMBASE, CINAHL, and MEDLINE) through February 2013 to identify studies. We also
reviewed reference lists of included articles to identify any missing sources.
Study Selection: Inclusion criteria required identification of sexual violence among refugees and
IDPs or those displaced by conflict in complex humanitarian settings. Studies were excluded if
they did not provide female sexual violence prevalence, or that included only single case reports,
anecdotes, and those that focused on displacement associated with natural disasters. After a
review of 1175 citations 19 unique studies were selected.
Data Extraction: Two reviewers worked independently to identify final selection and a third
reviewer adjudicated any differences. Descriptive and quantitative information was extracted;
prevalence estimates were synthesized. Heterogeneity was assessed using I2.
Main Outcomes: The main outcome of interest was sexual violence among female refugees and
IDPs in complex humanitarian settings.
Results: The prevalence of sexual violence was estimated at 21.4% (95% CI, 14.9-28.7;
I2=98.3%), using a random effects model. Statistical heterogeneity was noted with studies using
probability sampling designs reported lower prevalence of sexual violence (21.0%, 95% CI, 13.2-
30.1; I2=98.6%), compared to lower quality studies (21.7%, 95% CI, 11.5-34.2; I2=97.4%). We
could not rule out the presence of publication bias.
Conclusions: The findings suggest that approximately one in five refugees or displaced women
in complex humanitarian settings experienced sexual violence. However, this is a likely an
!
$!
underestimation of the true prevalence given the multiple existing barriers associated with
disclosure. The long-term health and social consequences of sexual violence for women and their
families necessitate strategies to improve identification of survivors of sexual violence and
increase prevention and response interventions in these complex settings.
!
%!
Background
War places civilians at increased risk of many forms of violence. The fear of violence can
promote forced and mass displacement. As of January 2012, almost 11 million people were
registered as refugees or internally displaced persons (IDPs), as estimated by UNHCR.1 During
transitions through conflict and displacement, refugees and internally displace persons (IDPs)
continue to live at heightened vulnerability to violence due to breakdown of family and social
structure, and changes to law enforcement and protective structures.2-4 Significant efforts have
been made to assess for, prevent and respond to gender-based violence (GBV) that occurs in these
settings2,5. However, GBV is broadly defined.6 For the purpose of this study, we focus on sexual
violence as defined by the US Center for Disease Control7:
…any nonconsensual completed or attempted contact (between the penis and the vulva
or the penis and the anus involving penetration, however slight), nonconsensual
contact between the mouth and the penis, vulva, or anus; nonconsensual penetration of
the anal or genital opening of another person by a hand, finger, or other object;
nonconsensual intentional touching, either directly or through the clothing, of the
genitalia, anus, groin, breast, inner thigh, or buttocks; or nonconsensual non-contact
acts of a sexual nature such as voyeurism and verbal or behavioral sexual harassment.
All the above acts also qualify as sexual violence if they are committed against
someone who is unable to consent or refuse…
Displaced women and girls are vulnerable to a range of sexual violence including forced sex/rape,
sexual abuse by an intimate partner, child sexual abuse, coerced sex, and sex trafficking in
conflict and humanitarian settings.8 Many studies have focused on the issue of rape as a weapon
of war, leading to assumptions that armed actors and military personnel are the main perpetrators
of sexual violence.9 Other perpetrators, however, may also include family members, NGO and
humanitarian workers, trusted individuals, or strangers who take advantage of heightened
!
&!
vulnerability.10,11 As a result, women and girls who experience sexual violence may experience a
range of long lasting physical,12,13 reproductive,14-16 and mental health consequences of sexual
violence.13.
Prevention and response to sexual violence in humanitarian settings focuses on three main areas
of medical and reproductive care, psychosocial support, and protection.5,17-23 To ensure quality
and coverage of these services, however, donors and humanitarian organizations must make
evidence-informed decisions with inputs related to level of need, intervention costs, and other
priority needs of the displaced population.24 To this end, response efforts for sexual violence
have been hampered by the lack of an adequate epidemiological understanding of the true
estimates of sexual violence among refugees and IDPs. The available literature is scant and
unreliable.25 The little research that is available has reported varying prevalence estimates, some
with extremely high or low figures,26 leading to concerns about inappropriate estimations of both
the true magnitude of sexual violence and the contexts in which sexual violence occurs.10,27
Differences in sampling techniques, definitions and recall periods of sexual violence, ethical
considerations, or the challenging nature of conducting research in complex humanitarian
emergencies may partially explain such differences in estimates.27
Though there have been systematic reviews of gender-based violence26 and sexual violence,9 to
our knowledge, there has been no published meta-analysis of prevalence of sexual violence
among displaced populations that has attempted to combine and contrast the estimates from
across these different studies. Our objective is to estimate the prevalence of sexual violence
among refugees and displaced populations in complex humanitarian emergencies.
!
'!
Methods
Literature Search
The meta-analysis was conducted according to a pre-specified protocol available from the
investigators. We conducted an initial search on MEDLINE, EMBASE and CINAHL in
November 2010 using an enhanced filter in consultation with an information specialist. Details of
our search strategy and terms are presented in the Online Supplement (Appendix 1). The search
was optimized for sensitivity and specificity through key articles identified by experts. At the
time of manuscript preparation, we updated the search on February 2013 and also evaluated the
bibliographies of included studies for relevant publications.
We restricted our searches to studies published in English and developed a search strategy for
MEDLINE based on medical subject headings (MeSH) terms and text words of key articles that
we identified a priori. Studies were not included or excluded on the basis of design but were
required to report the study design and methodology for inclusion in this analysis. Inclusion
criteria required the description of an evaluation of a screening tool, strategy, survey, or program
to identify sexual violence among refugees and IDPs or those displaced by conflict in complex
humanitarian settings. We excluded studies that did not provide prevalence of sexual violence for
females, or that included only single case reports, anecdotes, and those that focused on
displacement associated with natural disasters. We excluded studies that focused on female
genital mutilation as a form of sexual violence. We excluded studies in which it was unclear as to
whether the study population was migrant or refugees or if the results were not stratified on the
basis of migrant or refugee/IDP status. Two reviewers worked independently and in duplicate, to
review titles, abstracts, and full text versions of identified reports. A third reviewer met to discuss
and to adjudicate differences.
Data Abstraction
!
(!
Following the search, duplicate publications were removed. Preliminary screening included a
review of all titles and abstracts of identified studies from our searches, excluding those that
failed to meet selection criteria. The remaining articles underwent full-text evaluation for
inclusion eligibility. Data were abstracted from identified studies that reported the outcome
measures of female-targeted sexual violence, prevalence estimates, and other associations.
Study Characteristics
We extracted information related to the study characteristics, including country where the study
was conducted, country of origin of study population, participant age range, proportion of female
participants within the study sample, and total sample size of female participants. We also
extracted information that described the study design, including design sampling method, whether
the target sample size was reached according to the study authors, non-response percentage.
Characteristics of the instrument used to assess sexual violence were also collected, including if
pilot testing of questionnaires was performed prior to data collection, the type of instrument(s)
used, and validation measures (i.e. reporting of internal consistency, sensitivity, and specificity).
Risk of bias
In order to determine the risk of bias of studies, we evaluated whether reliability had been
assessed and whether authors evaluated construct validity of the screening instruments. We also
determined whether sampling was convenience or probability-based. We did not conduct
quantitative tests for publication bias but assessed this qualitatively when relevant.
Outcome Measure
Sexual violence outcomes included reported rape, molestation, sexual abuse, gang rape, marital
rape, sexual violence related to exploitation, and sexual harassment, as reported by the authors.
Though recognizing the importance of GBV on the health and well-being of displaced persons,
!
)!
we focused the search on sexual violence among female refugees and IDPs, as opposed to GBV,
given the broad definition,6 interpretation, and variable measurement of GBV in these settings.2
Data Synthesis
Statistical analysis was conducted using StatsDirect (version 2.7.9). Statistical heterogeneity was
tested and prevalence proportions were pooled using a fixed-effect model if heterogeneity was
limited; a random-effect model was used when there was a significant heterogeneity among the
studies.28 Since heterogeneity was anticipated in these studies we considered the more
conservative random effects model as a more reliable estimate of the prevalence. Results from the
fixed effects model were also examined to test the robustness of our results. To maintain
similarity of sampling designs across the studies, we conducted sensitivity analyses to determine
the robustness of the effect size when studies that had a defined sampling design were analyzed
separately from those studies that reported non-probability based sampling methods. A
conventional level of p<0.05 was utilized to assess significance. Our study was reported out
according to the MOOSE Reporting instrument.29 The protocol is available on request from the
corresponding author.
!
*!
Results
Search results
After a review of 1175 citations, we selected 19 unique studies that reported on prevalence of
sexual violence among female refugees and IDPs in the setting of complex humanitarian
emergencies for inclusion.3,4,30-46 The selection of studies included in our review is summarized
in Figure 1.
Study characteristics
Characteristics of studies are summarized in Table 1. The 19 studies enrolled a total of 8398
participants. Participants of the studies were either refugees or IDPs from 14 different countries
of origin that were affected by conflict and 14 different countries where the studies were
conducted. The proportion of female participants ranged from 26.5% to 100%. The study sample
sizes ranged from 34 to 991. There was significant variability of the age range; the widest
reported age range was 11-70 years of age.
Design and risk of bias results
Table 2 summarizes the study designs and the survey instruments used to assess sexual violence
among female refugees and IDPs, as reported by the study authors. Among the nineteen selected
studies, 11 studies utilized probability based random sampling methods3,4,30,34-36,39,41,43,44,46 and
eight utilized non-probability based sampling methods.31-33,37,38,40,42,45 Definitions of sexual
violence ranged from “improper sexual acts of any kind”45 to narrowly specified acts of sexual
violence such as coerced penetration.4,34,36 Recall periods of sexual violence varied from 6
months41 to lifetime.4,31,34,39 Many of the studies did not develop the sample size with the
appropriate effect size to measure the prevalence of sexual violence as the principal aim of the
study. For example, several studies used a subsample of a larger study population to estimate the
sexual violence prevalence. 30-33,39,40,42-46 Six studies reported that they had reached the targeted
!
"+!
sample size.33,35,39,41,43,46 Another three studies reported that targeted sample size was not
reached.4,34,36 Non-response rate was reported in nine studies.4,31,33-36,40,41,46 Five studies had
reported that pilot testing of the survey instruments was carried out prior to data
collection.4,33,34,36,46
The survey instruments used to assess for sexual violence the Harvard Trauma Questionnaire
(HTQ),47,48, the Abuse Assessment Screen (AAS)49 and survey questionnaires (validation of
survey questionnaires were not reported). The Internal consistency (Cronbach’s !) of the
instruments used was reported in the Roberts study43 that used HTQ. None of the studies reported
sensitivity and specificity of the instrument used to assess for sexual violence. We could not rule
out the presence of publication bias among the included studies.
Prevalence proportion for sexual violence
The estimated prevalence of sexual violence among the 19 selected studies3,4,30-46 was 21.4%
(1521/8398; 95% CI, 14.9-28.7; I2=98.3), using the random effects statistical model (Figure 2).
A sensitivity analysis was performed to compare the prevalence proportion of sexual violence of
the studies that used probability based random sampling to the studies that used non-probability
based sampling (Appendix 2). Compared to the primary analysis, the 11 studies that utilized
probability based random sampling methodologies resulted in an estimated 21.0% prevalence of
sexual violence (961/6265; 95% CI, 13.2-30.1; I2=98.6%) using the random effects model. The
eight studies that used non-probability based sampling methodologies produced an estimated
prevalence of sexual violence of 21.7% (560/2133; 95% CI, 11.5-34.2; I2=97.4%) using the
random effects model. Another sensitivity analysis was done to compare studies that focused on
sexual violence as the primary objective compared to studies that was not dedicated to sexual
violence as the primary objective (Appendix 2). Compared to the primary analysis, the 9 studies
that focused on sexual violence as the primary objective yielded in an estimated 20.7%
(911/4476; 95% CI, 13.0-29.6; I2= 97.8%) compared to the 10 studies with other primary
!
""!
objectives with an estimated 22.2% (610/3922; 95% CI, 11.8-34.7; I2= 98.6%) prevalence of
sexual violence using random effects model.
Discussion
In the present meta-analysis of 19 studies, we found the prevalence of sexual violence among
female refugees and internally displaced persons across 14 countries affected by conflict to be
21.4%. The quantification of sexual violence among female refugees and IDPs in complex
humanitarian emergencies is challenging.
Sexual violence is often under-reported. The social stigma associated with rape, shame and fear
of reprisal are significant deterrents for survivors of sexual violence to report their traumatic
experiences. Compounding social stigma is an often inadequate justice system response that fails
to arrest or prosecute perpetrators, a law enforcement system that often mistreats and further
victimizes survivors of sexual violence, and a lack of capacity of service providers across
multiple systems to receive and give adequate attention to the various and complex needs of
women who have been raped.2 The cumulative effect is an inhospitable climate for survivors to
come forth to disclose their experience and to seek help. The negative health impacts of the
experience of sexual violence are significant and long-term and may, include serious physical
injuries, sexually transmitted infections and HIV infections,16,50 fistulas and chronic pain,51,52
unwanted pregnancies,53 and a myriad of psychological health consequences including
suicide.43,45,54,55
The body of research aimed at understanding the prevalence of sexual violence among refugees
and displaced populations in conflicts is exponentially more difficult to elucidate given the
context and the sensitive subject matter. Not only is there difficulty in reaching the affected
populations, but investigators also employ varying definitions and research methodologies to
estimate prevalence of sexual violence. Even when taking into account the methodological issues,
!
"#!
our finding provides evidence of the need for concerted action to address sexual violence among
refugees and displaced populations in conflict settings. The estimated lifetime prevalence of
sexual violence against women in non-conflict settings ranges from 10%-50% depending on the
country and who is identified as the perpetrator (e.g. intimate partner vs. stranger).13,56 In many
countries where these estimates are collected in non-conflict setting, basic justice and law
enforcement systems are in place. Situations of conflict and displacement may exacerbate
existing gender based violence in families and communities and present new forms of violence
(e.g. sexual slavery) against women and girls.53 Hence, the prevalence of sexual violence of
21.4% among refugees and IDPs still likely underestimates the true prevalence as many
incidences of sexual violence in the setting of complex humanitarian emergencies, such as
conflict, go unreported.
Similar to the recent systematic review by Stark and Ager,26 which focused on the prevalence of
GBV in complex emergencies, the results of this systematic review calls for a need for more
uniform methods and common definitions in future research. Greater uniformity would yield
deeper understanding of sexual violence among refugee and displaced populations. Adopting
more consistent approaches, moreover, could potentially identify greater incidence of sexual
violence.
In addition to the importance of determining prevalence, there is also a critical need to understand
who the range of perpetrators may be, as well as the physical locations and settings in which
sexual violence is likely to occur. More vigorous action is needed to prevent and to respond to
sexual violence among refugees and displaced populations, identify methods to assist survivors,
and hold perpetrators accountable. Potential steps include strengthening procedures for
identifying survivors, such as the integration of routine GBV screening inquiry for women and
girls in protection and health programs using a validated measure with trained providers;
expanding prevention efforts using social norms perspectives, providing response services such as
!
"$!
psychosocial services and livelihood programs for survivors;23 building competency, compassion
and collaboration among protection officers, health and other service providers, and police; and
providing resources and political will to investigate and prosecute perpetrators.
Limitations
Our analysis has limitations. Our search focused on peer-reviewed publications and searches of
secondary sources. Our search may have exclude non-English articles and non peer-reviewed
reports by humanitarian organizations. Although sexual violence is a global phenomenon, the
search identified studies of refugees or displaced persons from predominantly African countries
and thus limits the generalizability of our findings to other contexts. The results may be
influenced by bias inherent in individual studies, particularly including social desirability bias
that often accompanies self-reported responses to sensitive questions.57 Some studies evaluated
violence and traumatic events that occurred in very distant past or did not report the time of the
event, thus it is important to note that we cannot deduce estimates of temporality. Since there are
no uniformly accepted measures of risk of bias assessments of cross-sectional studies, we
developed our own assessment scale for this context. Additionally, some of the select 19 studies
included in the final quantitative synthesis were not dedicated to assessing sexual violence but
rather on other related topics in which sexual violence questions were included in the survey. We
have performed a sensitivity analysis to compare the studies with the primary objective to
measure sexual violence compared to studies with other primary objectives and we found no
significant difference. The ideal study focused on assessing prevalence of sexual violence should
include survey questions that are based on specific types and acts of sexual violence and
delivered by trained interviewers in private confidential settings. As some of the studies were not
dedicated studies focused on sexual violence and many of the studies did not detail how the
questionnaires were administered, it is unclear how the final prevalence proportions reported
were affected.
!
"%!
While this analysis focused on sexual violence among females, sexual violence among males
exists and warrants future attention. Emerging reports have documented sexual and gender-based
violence among male refugee and IDP populations25,35,58,59 and response to prevent and address
health and social outcomes among these men have been emphasized as a critical component for
an inclusive and comprehensive response to sexual and gender-based violence among displaced
populations.60 We acknowledge this as an important area of study and believe a separate study to
assess the prevalence of sexual violence against men in displaced setting is warranted to give due
justice and attention to the issue.
Conclusions
The findings suggest that approximately one in five refugees or displaced women in complex
humanitarian settings experienced sexual violence. However, this is a likely an underestimation
of the true prevalence given the multiple existing barriers associated with disclosure. The long-
term health and social consequences of sexual violence for women and their families necessitate
strategies to improve identification of survivors of sexual violence and increase prevention and
response interventions in these complex settings.
Acknowledgments:
Thanks are due to the Center for Public Health and Human Rights for its ongoing support of this
research. This project was funded as a gift of the U.S. Government (U.S Department of State,
Bureau of Population, Refugees, and Migration). The funders had no role in study design, data
collection and analysis, decision to publish, or preparation of the manuscript. The authors have
declared that no competing interests exist.
!
"&!
References:
1. UNHCR. UNHCR Global Appeal 2013 Update: Populations of Concern. Geneva: United Nations
High Commissioner for Refugees;2013.
2. United nations High Commissioner for Refugees. Sexual and gender-based violence against
refugees, returnees, and internally displaced persons. Geneva May 2003.
3. Hagan J, Rymond-Richmond W, Palloni A. Racial targeting of sexual violence in Darfur. Am J
Public Health. Aug 2009;99(8):1386-1392.
4. Amowitz LL, Reis C, Lyons KH, et al. Prevalence of war-related sexual violence and other
human rights abuses among internally displaced persons in Sierra Leone. JAMA. Jan 23-30
2002;287(4):513-521.
5. Interagency Standing Committee. Guidelines for Gender-based Violence Interventions in
Humanitarian Settings. Geneva: The IASC Taskforce on Gender in Humanitarian
Assistance,;2005.
6. United Nations General Assembly. Declaration on the Elimination of Violence Against Women.
In: United Nations, ed. A/RES/48/104. Vol 48/104. Geneva1993.
7. Basile K, Saltzman L. Sexual violence surveillance: uniform definitions and recommended data
elements. Atlanta: Centers for Disease Control and Prevention, National Center for Injury
Prevention and Control,;2002.
8. Ward J, Vann B. Gender-based violence in refugee settings. Lancet. Dec 2002;360 Suppl:s13-14.
9. Hynes M, Lopes Cardozo B. Sexual violence against refugee women. Journal of women's health
& gender-based medicine. Oct 2000;9(8):819-823.
10. Human Security Research Group. Human Security Report 2012: Sexual Violence, Education, and
War: Beyond the Mainstream Narrative. Vancouver: Simon Fraser Univ., ;2012.
11. Wirtz A, Glass N, Pham K, et al. Development of a screening tool to identify female survivors of
gender-based violence in a humanitarian setting: qualitative evidence base from research among
refugees in Ethiopia. Conflict and Health. 2013;(Under review).
12. Campbell J, Jones AS, Dienemann J, et al. Intimate partner violence and physical health
consequences. Archives of internal medicine. May 27 2002;162(10):1157-1163.
13. Ellsberg M, Jansen HA, Heise L, Watts CH, Garcia-Moreno C. Intimate partner violence and
women's physical and mental health in the WHO multi-country study on women's health and
domestic violence: an observational study. Lancet. Apr 5 2008;371(9619):1165-1172.
14. Draughton. Sexual assault injuries and increased risk of HIV transmission. Advanced emergency
nursing journal. 2012;34(1):82-87.
15. McLean I, Roberts SA, White C, Paul S. Female genital injuries resulting from consensual and
non-consensual vaginal intercourse. Forensic science international. Jan 30 2011;204(1-3):27-33.
16. M. Olynik MGP, M.E. Cannon, G. Lachapelle. Temporal Variability of GPS Error Sources and
Their Effect on Relative Positioning Accuracy. Paper presented at: 2002 National Technical
Meeting of The Institute of Navigation2002; San Diego, CA.
17. Talen E, Shah S. Neighborhood evaluation using GIS - An exploratory study. Environment and
Behavior. Sep 2007;39(5):583-615.
18. Hankins CA, Friedman SR, Zafar T, Strathdee SA. Transmission and prevention of HIV and
sexually transmitted infections in war settings: implications for current and future armed
conflicts. AIDS. Nov 22 2002;16(17):2245-2252.
19. Hawkins SR, Clinton-Sherrod AM, Irvin N, Hart L, Russell SJ. Logic models as a tool for sexual
violence prevention program development. Health Promot Pract. Jan 2009;10(1 Suppl):29S-37S.
20. Pronyk PM, Hargreaves JR, Kim JC, et al. Effect of a structural intervention for the prevention of
intimate-partner violence and HIV in rural South Africa: a cluster randomised trial. The
Lancet.368(9551):1973-1983.
!
"'!
21. Glass N, Ramazani P, Tosha M, Mpanano M, Cinyabuguma M. A Congolese-US participatory
action research partnership to rebuild the lives of rape survivors and their families in eastern
Democratic Republic of Congo. Global public health. 2012;7(2):184-195.
22. Kalichman SC, Simbayi LC, Cloete A, et al. Integrated gender-based violence and HIV Risk
reduction intervention for South African men: results of a quasi-experimental field trial. Prev Sci.
Sep 2009;10(3):260-269.
23. Kohli A, Makambo MT, Ramazani P, et al. A Congolese community-based health program for
survivors of sexual violence. Conflict and health. 2012;6(1):6.
24. Leaning J, Spiegel P, Crisp J. Public health equity in refugee situations. Conflict and Health.
2011;5(1):6.
25. Leiby M. Digging in the Archives: The Promise and Perils of Primary Documents. Politics &
Society. March 1, 2009 2009;37(1):75-99.
26. Stark L, Ager A. A systematic review of prevalence studies of gender-based violence in complex
emergencies. Trauma, violence & abuse. Jul 2011;12(3):127-134.
27. Palermo T, Peterman A. Undercounting, overcounting and the longevity of flawed estimates:
statistics on sexual violence in conflict. Bulletin of the World Health Organization. Dec 1
2011;89(12):924-925.
28. Bradburn MJ, Deeks JJ, Berlin JA, Russell Localio A. Much ado about nothing: a comparison of
the performance of meta-analytical methods with rare events. Stat Med. Jan 15 2007;26(1):53-77.
29. Moher D, Liberati A, Tetzlaff J, Altman DG, The PG. Preferred Reporting Items for Systematic
Reviews and Meta-Analyses: The PRISMA Statement. PLoS Med. 2009;6(7):e1000097.
30. Amone-P'Olak K. Mental states of adolescents exposed to war in Uganda: finding appropriate
methods of rehabilitation. Torture. 2006;16(2):93-107.
31. Avdibegovic E, Sinanovic O. Consequences of domestic violence on women's mental health in
Bosnia and Herzegovina. Croat Med J. Oct 2006;47(5):730-741.
32. Betancourt TS, Borisova, II, de la Soudiere M, Williamson J. Sierra Leone's child soldiers: war
exposures and mental health problems by gender. J Adolesc Health. Jul 2011;49(1):21-28.
33. Hammoury N, Khawaja M. Screening for domestic violence during pregnancy in an antenatal
clinic in Lebanon. European Journal of Public Health. Dec 2007;17(6):605-606.
34. Hynes M, Robertson K, Ward J, Crouse C. A determination of the prevalence of gender-based
violence among conflict-affected populations in East Timor. Disasters. Sep 2004;28(3):294-321.
35. Johnson K, Asher J, Rosborough S, et al. Association of combatant status and sexual violence
with health and mental health outcomes in postconflict Liberia. JAMA. Aug 13 2008;300(6):676-
690.
36. Johnson K, Scott J, Rughita B, et al. Association of sexual violence and human rights violations
with physical and mental health in territories of the Eastern Democratic Republic of the Congo.
JAMA. Aug 4 2010;304(5):553-562.
37. Kerimova J, Posner SF, Brown YT, Hillis S, Meikle S, Duerr A. High prevalence of self-reported
forced sexual intercourse among internally displaced women in Azerbaijan. Am J Public Health.
Jul 2003;93(7):1067-1070.
38. Kinyanda E, Musisi S, Biryabarema C, et al. War related sexual violence and it's medical and
psychological consequences as seen in Kitgum, Northern Uganda: A cross-sectional study. BMC
Int Health Hum Rights. 2010;10:28.
39. Lopes Cardozo B, Vergara A, Agani F, Gotway CA. Mental health, social functioning, and
attitudes of Kosovar Albanians following the war in Kosovo. JAMA. Aug 2 2000;284(5):569-577.
40. McKelvey RS, Webb JA. A pilot study of abuse among Vietnamese Amerasians. Child Abuse
Negl. May 1995;19(5):545-553.
41. Parmar P, Agrawal P, Greenough PG, Goyal R, Kayden S. Sexual violence among host and
refugee population in Djohong District, Eastern Cameroon. Glob Public Health. 2012;7(9):974-
994.
!
"(!
42. Petersen HD, Worm L, Olsen MZ, Ussing B, Hartling OJ. Human rights violations in
Burma/Myanmar. A two year follow-up examination. Dan Med Bull. Nov 2000;47(5):359-363.
43. Roberts B, Ocaka KF, Browne J, Oyok T, Sondorp E. Factors associated with post-traumatic
stress disorder and depression amongst internally displaced persons in northern Uganda. BMC
Psychiatry. May 19 2008;8.
44. Swiss S, Jennings PJ, Aryee GV, et al. Violence against women during the Liberian civil conflict.
JAMA. Feb 25 1998;279(8):625-629.
45. Usta J, Farver JA, Zein L. Women, war, and violence: surviving the experience. J Womens
Health (Larchmt). Jun 2008;17(5):793-804.
46. Vinck P, Pham PN. Association of exposure to violence and potential traumatic events with self-
reported physical and mental health status in the Central African Republic. JAMA. Aug 4
2010;304(5):544-552.
47. Shoeb M, Weinstein H, Mollica R. The Harvard trauma questionnaire: adapting a cross-cultural
instrument for measuring torture, trauma and posttraumatic stress disorder in Iraqi refugees. Int J
Soc Psychiatry. Sep 2007;53(5):447-463.
48. Mollica RF, Caspi-Yavin Y, Bollini P, Truong T, Tor S, Lavelle J. The Harvard Trauma
Questionnaire. Validating a cross-cultural instrument for measuring torture, trauma, and
posttraumatic stress disorder in Indochinese refugees. J Nerv Ment Dis. Feb 1992;180(2):111-
116.
49. Laughon K, Renker P, Glass N, Parker B. Revision of the Abuse Assessment Screen to address
nonlethal strangulation. J Obstet Gynecol Neonatal Nurs. Jul-Aug 2008;37(4):502-507.
50. Campbell JC, Lucea MB, Stockman JK, Draughon JE. Forced Sex and HIV Risk in Violent
Relationships. American journal of reproductive immunology. Oct 16 2012.
51. Djuknic GM, Richton RE. Geolocation and assisted GPS. Computer. Feb 2001;34(2):123-125.
52. Longombe AO, Claude KM, Ruminjo J. Fistula and traumatic genital injury from sexual violence
in a conflict setting in Eastern Congo: case studies. Reproductive health matters. May
2008;16(31):132-141.
53. García-Moreno C, Jansen H, M. E, Heise L, Watts C. WHO multi-country study on women’s
health and domestic violence against women : initial results on prevalence, health outcomes and
women’s responses. World Health Organization: World Health Organization;2005.
54. Keller A, Lhewa D, Rosenfeld B, et al. Traumatic experiences and psychological distress in an
urban refugee population seeking treatment services. J Nerv Ment Dis. Mar 2006;194(3):188-194.
55. Wenzel T, Rushiti F, Aghani F, Diaconu G, Maxhuni B, Zitterl W. Suicidal ideation, post-
traumatic stress and suicide statistics in Kosovo. An analysis five years after the war. Suicidal
ideation in Kosovo. Torture. 2009;19(3):238-247.
56. Tjaden P, Thoennes N. Prevalence, Incidence, and Consequences of Violence Against Women:
Findings From the National Violence Against Women Survey. Rockville: CDC,;2000.
57. Vu A, Tran N, Pham K, Ahmed S. Reliability of the Marlowe-Crowne social desirability scale in
Ethiopia, Kenya, Mozambique, and Uganda. BMC Med Res Methodol. 2011;11:162.
58. Christian M, Safari O, Ramazani P, Burnham G, Glass N. Sexual and gender based violence
against men in the Democratic Republic of Congo: effects on survivors, their families and the
community. Medicine, conflict, and survival. Oct-Dec 2011;27(4):227-246.
59. UN OCHA. Discussion Paper 2: The nature, scope, and motivation for sexual violence against
men and boys in armed conflict. Paper presented at: Use of Sexual Violence in Armed Conflict:
Identifying gaps in Research to Inform More Effective Interventions UN OCHA Research
Meeting – 26 June 20082008.
60. United Nations High Commissioner for Refugees. Working with men and boy survivors of sexual
and gender-based violence in forced displacement. Geneva: UNHCR,;2012.
!
")!
Table 1: Study characteristics
Study Author
(Year of study)
Country
of Study
Country of Origin of
refugee/displaced
Age Range
(years)
Proportion of
Female
Participants (%)
No. of
Females
in Sample
(N)
McKelvey (1995) Philippines Vietnam 18.0 - 25.0 33.0 34
Swiss (1998) Liberia Liberia Not reported 100.0 205
Petersen (2000) Thailand Burma/Myanmar 11.0 - 70.0 37.0 48
Cardozo (2000) Kosovo Kosovo Not reported 62.3 825
Amowitz (2002) Sierra Leone Sierra Leone 14.0 - 80.0 100.0 991
Kerimova (2003) Azerbaijan Azerbaijan 33.5 * 100.0 457
Hynes (2004) E. Timor E. Timor 18.0 - 49.0 100.0 287
Avdibegovi! (2006) Bosnia and Herzegovina Bosnia and Herzegovina 43.0 * 100.0 283
Amone-P'Olak
(2006)
Uganda Uganda 12.0 - 19.0 26.5 78
Hammoury (2007)
Lebanon Palestine 28.0 * 100.0 349
Roberts (2008) Uganda Uganda 35.3 * 60.1 727
Johnson (2008) Liberia Liberia 40.2 - 42.4 52.8 880
Usta (2008) Lebanon Lebanon 15.0 - 72.0 100.0 310
Hagan (2009)
Chad Sudan 37.1 * 60.0 559
Kinyanda (2010) Uganda Uganda 24.0 * 70.5 573
Vinck (2010)
Central African
Republic
Central African
Republic
36.4 * 49.8 936
Johnson (2010) DRC DRC 38.2 - 42.0 59.4 586
Betancourt (2011) Sierra Leon Sierra Leon 16.2 * 28.9 79
Parmar (2012) Eastern Cameroon
Central African
Republic 35.1 * 100.0 191
* Age range not available. Mean age reported instead
!
"*!
Table 2: Reported Study Design and Study Instrument Used to Assess Sexual Violence
Study Author
(Year of study)
Study Design SV Assessment Instrument
Probability
vs.
Non-
Probability
Sample
Size
Adequacy
Reached
Non-
response
Pilot
Testing
Type of
Instrument Used
Sensitivity/
Specificity
McKelvey (1995) Non-Probability Not reported 5.9% Not reported Survey/Questionnaire Not reported
Swiss (1998) Probability Not reported Not reported No Survey/Questionnaire Not reported
Petersen (2000)
Non-Probability
Not reported
Not reported
Not reported
Survey/Questionnaire
Not reported
Cardozo (2000) Probability Yes Not reported No HTQ Not reported
Amowitz (2002) Probability No 5.0% Yes Survey/Questionnaire Not reported
Kerimova (2003) Non-Probability Not reported Not reported No Survey/Questionnaire Not reported
Hynes (2004)
Probability
No
26.0%
Yes
Survey/Questionnaire
Not reported
Avdibegovi! (2006) Non-Probability Not reported 0.0% Not reported Survey/Questionnaire Not reported
Amone-P'Olak (2006) Probability Not reported Not reported Not reported Survey/Questionnaire Not reported
Hammoury (2007) Non-Probability Yes 0.6% Yes AAS Not reported
Roberts (2008)
Probability
Yes
Not reported
Not reported
HTQ
Not reported
Johnson (2008)
Probability
Yes
1.8%
Not reported
Survey/Questionnaire
Not reported
Usta (2008) Non-Probability Not reported Not reported No Survey/Questionnaire Not reported
Hagan (2009) Probability Not reported Not reported No Survey/Questionnaire Not reported
Kinyanda (2010) Non-Probability Not reported Not reported No Survey/Questionnaire Not reported
Vinck (2010)
Probability
Yes
5.0%
Yes
Survey/Questionnaire
Not reported
Johnson (2010) Probability No 1.1% Yes Survey/Questionnaire Not reported
Betancourt (2011) Non-Probability Not reported Not reported No Survey/Questionnaire Not reported
Parmar (2012)
Probability
Yes
0.5%
No
Survey/Questionnaire
Not reported
*HTQ- Harvard Trauma Questionnaire **AAS- Abuse Assessment Screen
Articles identified through literature search
n = 1175
• Electronic databases = 1088
• Identified through reference searches = 20
• Identified through other sources = 66
Screening of Titles and Abstracts
n = 1175
Excluded n = 1007
• Not original study n = 143
• Duplicates n = 14
• Does not pertain to female
sexual violence n = 542
• Does not include population
of interest n = 160
• Single case report n = 5
• Other reviews,
commentaries, letters not
relevant to population or
study n = 143
Full-text articles reviewed for eligibility
n = 168
Excluded n = 149
• Not original study n = 35
• Does not include sample
prevalence for female sexual
violence n = 4
• Does not pertain to female
sexual violence n = 28
• Does not include population
of interest n = 65
• Pure qualitative study n = 2
• Other reviews,
commentaries, letters not
relevant to population or
study n = 15
Studies included in meta-analysis and
sensitivity analysis
n = 19
Proportion meta-analysis plot [random effects]
0.0 0.2 0.4 0.6 0.8 1.0
combined 0.212 (0.147, 0.286)
Parmar (2012) 0.408 (0.338, 0.482 )
Betancourt (2011) 0.443 (0.331, 0.559 )
Johnson (2010) 0.382 (0.343, 0.423)
Vinck (2010) 0.059 (0.045, 0.076 )
Kinyanda (2010) 0.286 (0.250, 0.325)
Hagan (200 9) 0.070 (0.050, 0.094)
Usta (2008) 0.016 (0.005, 0.037)
Johnson (2008) 0.163 (0.139, 0.189)
Roberts (2008) 0.180 (0.153, 0.210 )
Hammoury (2007) 0.235 (0.194, 0.280 )
Amone-P'Olak (2006) 0.833 (0.732, 0.908)
Avdibegovic (2006) 0.435 (0.376, 0.495 )
Hynes (2004) 0.226 (0.179, 0.279 )
Kerimova (2003) 0.295 (0.254, 0.340)
Amowitz (2002) 0.095 (0.077, 0.115)
Cardozo (2000) 0.044 (0.034, 0.057)
Petersen (2000) 0.063 (0.013, 0.172 )
Swiss (1998) 0.151 (0.105, 0.208)
McKelvey (1995) 0.088 (0.019, 0.237)
proportion (95% confidence interval)
No. of Females in
Sample (N)!
No. of Females Reporting
Sexual Violence (n)!
McKelvey (1995)!
34!3!
Swiss (1998)!
205!31!
Petersen (2000)!
48!3!
Cardozo (2000)!
825!36!
Amowitz (2002)!
991!94!
Kerimova (2003)!
457!135!
Hynes (2004)!
287!65!
Avdibegović (2006)!
283!123!
Amone-P'Olak (2006) !
78!65!
Hammoury (2007)!
392!92!
Roberts (2008)!
727!131!
Johnson (2008)!
880!143!
Usta (2008)!
310!5!
Hagan (2009)!
559!39!
Kinyanda (2010)!
573!164!
Vinck (2010)!
936!55!
Johnson (2010)!
586!224!
Betancourt (2011)!
79!35!
Parmar (2012)!
191!78!
combined!
8398!1521!
Figure'2:'Meta-analysis'plot'
Appendix 1. Search terms
PUBMED/MEDLINE
EMBASE
CINAHL
Questionnaires; Tool; Evaluation; Measurement;
Diagnosis; Diagnostic tool; Assessment
Questionnaires;
Evaluation; Measurement;
Screening
Child Abuse; Dowry-related violence; Rape;
Marital rape; Circumcision, female; Female genital
mutilation; Female genital cutting; Non-spousal
violence; Sexual violence related to exploitation;
Sexual Harassment; Sexual harassment;
Trafficking; Forced prostitution; Systematic rape;
Sexual slavery
Partner violence; Sexual
abuse; Rape; Child sexual
abuse; Prostitution
Intimate partner violence;
Sexual abuse; Rape; Child
sexual abuse;
Refugees; Emigrants; Immigrants; Transients; War
Refugees; War; Conflict
!
!
Appendix 2. Sensitivity analysis for reported prevalence of sexual violence
Types of studies included in
sensitivity analysis
Statistical
Model No. of Studies No. Events,
n/N
Prevalence
Proportion,
% (95% CI)
Probability based random sampling
Random 11 961/6265 21.0 (13.2-30.1)
Non-probability based sampling Random 8 560/2133 21.7 (11.5-34.2)
Studies with primary objective to
measure sexual violence
Random 9 911/4476 20.7 (13.0-29.6)
Studies with other objective but
reported on sexual violence
Random 10 610/3922 22.2 (11.8-34.7)
!
