Background
The current humanitarian crisis in Syria, which began during spring 2011, continues to displace Syrians across the country [
1‐
4]. The number of displaced Syrians who have fled to nearby Lebanon is now over 700,000 (October 2013) up from 48,000 just over a year ago (August 2012) [
4]. About 24% of these refugees are women between the ages of 18–59 [
4]. The mountainous border region of North Lebanon and the Bekaa Valley along Lebanon’s eastern border have received the majority of refugees (34% and 29%, respectively) [
4].
North Lebanon and the Bekaa Valley are underserved regions with nearly 53% and 30%, respectively, living below the poverty line in 2008 [
5]. Additionally, based on a multidimensional poverty index, 49% of households in North Lebanon and 45% in the Bekaa Valley were deprived of health services in 2004 [
6]. With the influx of Syrian refugees, resources in these two regions are now even more strained.
Research shows that women may be more susceptible to poor reproductive health outcomes and violence, including sexual and gender-based violence (SGBV), in conflict settings and refugee crises [
7‐
13]. SGBV has been found to increase surrounding armed conflict, both as a weapon of war and as intimate partner violence, particularly in refugee settings, and reported SGBV prevalence varies greatly [
10,
12‐
15]. Violence against women (VAW) can have direct health consequences such as injury, psychological trauma or stress, and gynecologic problems [
8,
16]. Additionally, stress may be a mediator between exposure to violence in conflict settings and a variety of poor health outcomes among women [
15‐
17]. Numerous anecdotal reports describe VAW related to the Syrian conflict, but assessment of this critical issue is scarce [
18‐
21].
The relationship between conflict violence and reproductive health outcomes may be affected by a variety of factors, including: stress or anxiety [
17,
22]; socio-demographic indicators such as age, education, and marital status [
22,
23]; location in Lebanon, urban vs. rural status, and months as a refugee, which could be related to exposure to violence for geopolitical reasons and may also pertain to certain reproductive health outcomes based on healthcare access or local practices [
23]; food insecurity resulting from violence and displacement, which may contribute to poor reproductive health outcomes through behavioral, nutritional, or mental health pathways [
8,
16]; cigarette smoking [
16]; anemia [
10]; and hypertension [
24]. We assessed the potential confounding effects of such variables and adjusted our analyses accordingly.
The overall goal of this study was to increase understanding of reproductive health concerns in a conflict setting by assessing the experiences of displaced women in Lebanon who have recently fled from the conflict in Syria. This aligns with research priorities set by the Interagency-Working Group (IAWG) on Reproductive Health in Crisis and adds to the existing literature on violent conflict, SGBV, and refugee reproductive health by extending current research to a Middle Eastern setting that is experiencing ongoing conflict [
25].
Methods
Study participants and location
This needs assessment was conducted in Lebanon between June and August 2012, as refugee numbers were escalating in Lebanon one year after the Syrian conflict began. The study was carried out at three primary health clinics located in each of two regions of Lebanon: North Lebanon and the Bekaa Valley under the auspices and at the request of the United Nations Population Fund (UNFPA) Beirut, Lebanon office. Clinics were selected based on the number of displaced Syrian women attending per month (at least 100) and the provision of reproductive health services. In order to minimize bias related to clinic selection, a mix of clinics supported by government and non-government sources was chosen with all receiving some type of support from UNFPA. Three clinics were supported by a private foundation, two clinics were jointly run by the Lebanese Ministry of Public Health (MOPH) and a private foundation, and one clinic was run solely by MOPH. Since none of the local clinics had their own IRBS, ethics approval was sought and obtained from the Human Subjects Committee at Yale School of Public Health (YSPH) and UNFPA/Lebanon using standard procedures for written approval of study protocols and all ethical standards for human subjects research were adhered to throughout the study period.
A cross-sectional survey was carried out in the six primary healthcare clinics. We used a proportional sampling method, based on the number of Syrian women attending each clinic during the month prior to the study, to recruit at least 400 displaced Syrian women. All female, displaced Syrians presenting to these six clinics within the month of July 2012 were approached, screened for eligibility, and asked if they would like to participate until the target number was reached. Eligibility criteria included: ability to speak Arabic, identity as a Syrian national, arrival in Lebanon since the conflict in Syria began in March 2011, and age between 18 and 45 (inclusive). Once screened, women were escorted into a private room where an IRB-approved consent form was explained and signed prior to questionnaire administration.
Data collection
The interviewer-administered questionnaire was adapted from the “Gender-based Violence Tools Manual For Assessment & Program Design, Monitoring & Evaluation in Conflict-Affected Settings” and the “Reproductive Health Assessment Toolkit for Conflict-Affected Women” [
26,
27]. The questionnaire was designed in English, discussed with the stakeholders, translated into Arabic and pilot tested among Syrians in Lebanon, then administered in Arabic by trained female research assistants from the area. It addressed the following topics: 1) individual and displacement characteristics; 2) general health status; 3) reproductive history and current status; 4) pregnancy; 5) exposure to violence during the conflict, including sexual violence; and 6) help-seeking behaviors and stress. Of the 489 Syrian women approached to participate, nine did not meet eligibility criteria, 28 declined, and 452 (92.4%) completed the interview. Those who declined to participate cited the following reasons: not interested (28.6%), husband/family member would not allow (25.0%), ill/incapacitated (21.4%), fear (17.9%), and other reason or missing (7.1%). At the end of the interview, participants received a UNFPA “dignity kit” containing basic sanitation supplies and clothing to compensate for their time and were also given telephone numbers for agencies providing psychosocial and other resources for survivors of violence.
Data analysis
Bivariate associations were estimated using Pearson correlations and X
2 test. Risk factors associated with any of the outcomes of interest, and covariates associated with both risk factors and outcomes, at the level of p < 0.05 were retained in multivariate models. Multivariate logistic regression was used to examine relationships between independent variables (exposure to conflict violence and stress score) and health outcomes (gynecologic outcomes, self-rated health, and reproductive health services access).
A positive response (“1–2 times” or “frequently”) to any of the indicators of violence from an armed person since the conflict began (including: being slapped or hit; choked; beaten or kicked; threatened with a weapon; shot at or stabbed; detained against will; intentionally deprived of food, water, or sleep; emotional abuse or humiliation; deprived of money; or subjected to improper sexual behavior) was considered “exposure to violence,” which was coded as a binary variable. Stress was assessed using a 6-question subscale used previously by UNFPA among conflict-affected populations in Lebanon, with the addition of a question about child beating as an indicator of stress based on qualitative findings among Syrian women. Questions covered the following: feeling constantly tense, sick or tired, worried or concerned, irritable or in a bad mood, suffering from loss of sleep or sleep disorders, reduced ability to complete normal tasks, and beating or taking anger out on children. Principle-component analysis was conducted to determine if this subscale accurately reflected the construct of stress, and to create a stress score variable based on participant responses. The following variables were examined as potential confounders in the relationship between violence or stress and reproductive health outcomes using bivariate analysis; biologically plausible potential confounders were controlled for in multivariate analyses: age, education, marital status, region in Lebanon, clinic and clinic type (government funded or not), place of origin (urban versus rural), months in Lebanon, food insecurity indicators, cigarette smoking, anemia, and hypertension. Data were analyzed using Statistical Analysis Software (SAS) version 9.2, and principle-component analysis was carried out in IBM SPSS Statistics 21.
Discussion
Our findings indicate that Syrian women displaced to Lebanon experience various indicators of poor reproductive health, including: gynecologic conditions, pregnancy and delivery complications, and poor birth outcomes. High reported rates of menstrual irregularity, severe pelvic pain, and vaginal infections align with previous research on gynecologic outcomes in situations of violence or refugee settings [
9,
16]. Previous research also indicates that poor pregnancy outcomes, including low birthweight and preterm birth, may be related to refugee status, inadequate antenatal care, and economic hardship [
9,
10,
29,
30]. All of these factors were present among displaced Syrian women in our sample. In addition to poor reproductive health outcomes, many women rated their health as poor and this was statistically significantly associated with exposure to violence when mediated by stress. Many reported having chronic illnesses, including anemia and hypertension, which may be related to complications surrounding pregnancy and delivery [
15,
31,
32]. Food insecurity, identified among more than half of respondents, may be contributing to menstrual irregularity or to increased anemia.
Exposure to conflict-related violence, abuse, and/or sexual violence was reported by over 30% of women and many women reported multiple types of violence. While several cases of sexual violence perpetrated by armed people in Syria were reported, there may be underreporting of sexual violence due to shame or fear of stigmatization, despite our adherence to WHO guidelines for interviewing survivors of violence [
33]. Although WHO has no directly comparable prevalence data on conflict violence or non-partner sexual violence in the Eastern Mediterranean region, we can compare our findings to WHO data in other regions. South East Asia for example has a reported lifetime prevalence of non-partner sexual violence of 4.9%, while our study found a 3.1% prevalence of non-partner sexual violence during the conflict in Syria [
15]. Additionally, WHO found a 37.0% prevalence of intimate partner violence in the Eastern Mediterranean region, which suggests that the 30.8% prevalence of conflict violence that we found in our sample may be in addition to violence women are experiencing in the home [
15]. Multivariate analyses revealed significant positive associations between exposure to conflict violence and gynecologic conditions (menstrual irregularity, severe pelvic pain, and RTIs), which is consistent with existing literature in both refugee and non-refugee populations, and extends findings focused primarily on intimate partner violence to conflict-related violence. With the recent rise in conflict in the Middle East, it is useful to have a prevalence estimate of conflict violence experienced by women for the purposes of humanitarian response planning.
While the majority of those who experienced conflict violence reported suffering from psychological difficulties, only a small percent (9.0%) accessed mental health services. A previous study found that the majority of Lebanese women exposed to conflict violence and suffering from psychological distress received help from their families while few of them sought medical care [
13]. Mental health and psychosomatic conditions reported among survivors of violence could also be a risk factor for poor reproductive health outcomes [
14]. Our results suggest a mediating role for stress between exposure to violence and some health outcomes, highlighting the need to include mental health services with other health services for refugees. Many women reported health conditions potentially related to stress, including: nerve issues, depression, unusual pain and fatigue, loss of appetite or sleep, repeated vomiting, and migraines. Outreach to refugee women regarding conflict-related stress could help reduce the burden of mental health and reproductive health conditions, and may attenuate the effect of VAW on reproductive health.
Previous research shows that limited access or delayed entry to antenatal care is a key determinant of pregnancy outcomes in refugee populations, and there are often significant disparities in access to and use of antenatal care among refugee populations compared to non-refugee populations [
9]. Our study supports this finding and goes on to identify perceived barriers to access among a Middle Eastern population. The majority of Syrian refugee women had never visited an obstetrician-gynecologist except for pregnancy care, indicating low baseline rates of gynecological exams. While costs and long distances were the primary barriers reported, one unique barrier was lack of availability of a female gynecologist (as many women requested to be seen only by a female doctor). These findings indicate the need to increase awareness of the importance of antenatal care and increase availability of female physicians to provide these services in a culturally sensitive manner.
This study has several limitations. Although we selected a proportionally representative sample, results cannot be generalized to all Syrian refugee women in Lebanon, particularly to those who may not have access to clinics. Since this is an ongoing conflict setting and the number of Syrian refugees is increasing over time, it is possible that newer refugees may have different characteristics (e.g., age, SES, ethnic/religious affiliations, etc.) or potentially different needs. However, this study provides much-needed baseline data to assess the rapidly evolving situation. In addition, we relied on self-report, which may be subject to under- or over-reporting and is of particular concern for key reproductive health outcomes and sexual violence. Although we tried to ensure privacy, revealing such personal information during an interview may still be sensitive. Also, sexual violence questions were not asked using “event-based” items (e.g. forced to engage in a sexual act when participant did not want to), which may increase under-reporting [
34]. Finally, the survey location at health care centers poses a limitation on generalizability and prevalence estimates, as women presenting to these clinics may differ from the general population with respect to health status, behaviors, and knowledge of health services. Due to the fast-moving nature of this conflict setting, there is no data available on the total number of female Syrian refugees attending clinics in Lebanon during the time period of our study. UNHCR reports that 248,000 Syrian refugees have been assisted with primary healthcare to date (August 2013), indicating that over half of Syrian refugees have not attended a health clinic (though this estimate is not disaggregated by gender) [
4]. Despite limitations, this study provides important information about a vulnerable population that continues to grow as the conflict shows no signs of abating. In addition, information gained from this study may assist in planning for future humanitarian crises involving large numbers of displaced women and children.
Acknowledgements
This study was funded by the United Nations Population Fund (UNFPA) and the Okvuran Fund for International Support from Yale School of Public Health. Amelia Reese Masterson received the 2013 Lowell Levin Award for Excellence in Global Health at Yale School of Public Health. The study was conducted in cooperation with the UNFPA (Lebanon office), International Medical Corps (IMC), the Rafik Hariri Foundation, the Makassad Association, the Lebanese Ministry of Public Health (MOPH), and the Danish Refugee Council (DRC) in Lebanon.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
ARM, JU, and ASE designed study, conducted literature search and wrote manuscript. ARM conducted study and analyzed data in consultation with JU and ASE. JG contributed to literature search, manuscript, and interpretation of results. All authors read and approved the final manuscript.