Background: Intimate partner violence during pregnancy is a public health problem that can affect both maternal and fetal life. However, its prevalence and associated factors have not been well studied and understood in Ethiopia. Hence, this study was conducted to assess the individual and community-level factors associated with intimate partner violence during pregnancy in Gammo Goffa Zone, South Ethiopia. Methods: A community-based cross-sectional study was conducted among 1,535 randomly selected pregnant women from July to October 2020. Data were collected using an interviewer-administered, standardized WHO multi-country study questionnaire and analyzed using STATA 14. A two level mixed-effects logistic regression model was used to identify factors associated with intimate partner violence during pregnancy. Results: The prevalence of intimate partner violence during pregnancy was found to be 48% (95% CI: 45–50%). Factors affecting violence during pregnancy were identified at the community and individual levels. Access to health facilities (AOR = 0.61; 95% CI: 0.43, 0.85), women feeling isolated from the community (AOR= 1.96; 95% CI: 1.04, 3.69), and strict gender role differences (AOR= 1.45; 95% CI: 1.03, 2.04) were among higher-level factors found to be significantly associated with intimate partner violence during pregnancy. Low decision-making power was found to increase the odds of experiencing IPV during pregnancy (AOR= 2.51; 95% CI: 1.28, 4.92). Similarly, maternal education, maternal occupation, living with the partner’s family, current pregnancy intended by the partner, dowry payment, and presence of marital conflict were among the individual- level factors found to increase the odds of experiencing intimate partner violence during pregnancy. Conclusions: The prevalence of intimate partner violence during pregnancy was high in the study area. Both individual and community-level factors had significant implications on maternal health programs related to violence against women. Socio-demographic and socio-ecological characteristics were identified as associated factors. Since it is a multifaceted problem, special emphasis has to be given to multi-sectoral approaches involving all responsible bodies to mitigate the situation.
A community-based cross-sectional study was conducted in the Gammo Goffa Zone between July and October 2020. Gammo Goffa Zone is one of the 14 zones of the Southern Nations, Nationalities, and Peoples Regional (SNNPR) State of Ethiopia. Its capital, Arba Minch, is located 505 km south of Addis Ababa, and 275 km southwest of Hawassa, the capital city of the region. Administratively, the zone is subdivided into 15 rural districts designated as ‘Woredas’ and two town administrations. According to the population projection of Ethiopia for all regions at the woreda level from 2014–2017, the zone had a total population of 2,043,668 (25). In this cross-sectional study, pregnant women were the study population. The required sample size was determined using a single population proportion formula based on the following assumptions: a 35.6% prevalence of IPV during pregnancy in Ethiopia (5), a 95% level of the confidence interval, and a 4% degree of precision. Due to the multistage cluster sampling method used, a design effect of 2 was considered. Finally, 10% was added to the non-response rate. Accordingly, the final sample size was calculated to be 1,210. However, this study was a baseline survey of a cohort study to determine the effect of IPV during pregnancy on maternal and neonatal health outcomes, in which 1,535 pregnant women were followed up. Thus, to increase the precision of the estimates and the power of the study, the sample size was increased to 1,535. The sample size is adequate to identify factors associated with IPV during pregnancy. A multi-stage cluster sampling technique was employed to identify the study participants. Initially, the zone was stratified in to town administrations and rural districts. Then, in the first stage, by considering time and logistics, six districts were selected randomly. In the second stage, all the selected districts were stratified into urban and rural kebeles. A kebele is the smallest administrative unit (in the government structure) that is considered a cluster in this study. Then, 3 rural kebeles and 1 urban kebele were randomly selected from each selected district. In this zone, there were two town administrations (Arba Minch and Sawla) with 11 and 6 kebeles respectively, and all were purposefully included. A total of 41 clusters were selected randomly. Then, at the household level, an enumeration of pregnant women was conducted in the selected kebeles to fix a sampling frame. After identifying households with pregnant women, proportional to sample size allocations were employed. Finally, a simple random sampling was carried out to identify respondents from the selected households as a study unit (Figure 1). Schematic presentation of sampling procedure for the cross sectional study on IPV during pregnancy. The dependent variable for this study was IPV during the current pregnancy. IPV during the current pregnancy was defined as the experience of at least one act of any form of violence (psychological, physical, or sexual violence) by women perpetrated by their current or most recent partners, during the current pregnancy period. Psychological violence was measured as the experience of one or more acts or threats of acts, such as (a) being insulted, (b) being humiliated, (c) being intimidated, or d) threatening to hurt the study participant or someone the study candidate cares about (3). Physical violence was defined as the experience of one or more acts of physical aggression, such as (a) being slapped or having something thrown at her that could hurt her, (b) being pushed or shoved, (c) being hit with a fist or something else that could hurt; (d) being kicked, dragged, or beaten up; (e) being choked or burned on purpose, and/or (f) being threatened with, or actually having, a gun, a knife, or another weapon used on her by an intimate partner (3). Sexual violence was measured as the experience of one or more acts, such as a) being physically forced to have sexual intercourse, when she did not want to, b) having sexual intercourse because she was afraid of what her partner might do, and/or c) being forced to do something sexual that she found humiliating or degrading to her by an intimate partner (3). The independent variables were divided into two levels. Level-1 (lower-level variables), included individual and household characteristics, such as socio-demography, wealth index, reproductive and obstetric characteristics, women’s autonomy, and partnership-related variables. The wealth index was computed using principal component analysis (PCA). Level 2 (higher-level) variables included community and societal characteristics, such as place of residence, access to health facilities, and socio-ecological factors. The independent variables were selected based on their relationship with the dependent variable identified through reviewing existing literature (3, 15–17). In the previous studies, they were described as existing at one level; in this study, they are identified as variables operating at different levels. A pre-tested interviewer-administered structured questionnaire was adapted from the WHO multi-country study of the VAW questionnaire (3). The indicators for the wealth index were adapted from EDHS (26). The questionnaire was prepared in English and translated to the local language (Amharic), and back-translated to English by another person to ensure its consistency and accuracy. Health extension workers were recruited, trained and deployed for data collection. The data collection process was supervised by trained supervisors and principal investigators. The data collectors and supervisors were recruited based on their eloquence in local languages, qualifications, and experience in data collection. The WHO’s practical guide for researching VAW was adopted and used by the research team (27). Furthermore, we didn’t encounter any disruption during the study period due to COVID-19 because there was no strict lockdown/shutdown in Ethiopia and the disease incidence was very slow. After the data were coded and entered into EpiData v 3.1, were exported to STATA 14 for cleaning, editing, and analysis. Descriptive statistics were computed and presented. Socioeconomic quintiles were determined using principal component analysis (PCA). Since the occurrence of IPV during pregnancy is affected at different levels, a mixed-effects multilevel logistic regression model was employed. A bivariate analysis was done using cross-tabulation to test the association between IPV during pregnancy and independent variables. All variables having P < 0.25 were considered candidates for the final model. In this analysis, a two-level binary logistic regression model was used. The individual and family-level characteristics were considered as lower-level variables, and the community and societal characteristics were treated as higher-level variables. Generally, two models were estimated. These were the intercept-only model; an empty model, that contained no covariates, and a full model that included lower-level (level-1) and higher-level (level-2) variables. The goodness of fit of the multilevel model was tested by the log-likelihood ratio (LR) test. Multicollinearity between independent variables was assessed, using the variance inflation factor (VIF). The study was approved for scientific and ethical integrity by the Research and Ethical Review Committee (RER) of the School of Public Health, the Institutional Review Board (IRB) of the College of Health Sciences, Addis Ababa University (Protocol number: 106/19/SPH). Written informed consent was sought from every study participant. For women under the age of 18, consent was obtained from their parents. The study strictly followed the WHO guidelines on ethical issues related to violence research (28). All interviews were conducted in complete privacy. Data collectors were instructed to refer women with serious psychological distress to health facilities and act accordingly. After the completion of interviews, data collectors were observed for 14 days. The data collectors wore protective face masks. A reasonable physical distance was kept between interviewers and interviewees during data collection.
N/A
DIMA AI Care
