Study setting
The study is conducted at two non-governmental hospitals in Nepal, Dhulikhel Hospital (DH) and Kathmandu Medical College (KMC). DH is a community tertiary center situated in Dhulikhel, east of Kathmandu. KMC is located centrally in Kathmandu. At DH, routine antenatal care for low-risk women is provided by nurse-midwives. The rest of the women at DH are cared for by obstetricians. At KMC all antenatal care is provided by obstetricians.
Eligibility criteria and recruitment
All women attending routine antenatal care at DH or KMC, aged 18 or more, with a pregnancy between 12 – 22 weeks gestational age are invited to participate by study staff. Initially, women are informed that the study seeks to improve reproductive health and involves completing two questionnaires and watching a video twice. Women expressing interest to participate further are then taken to a private room where it can be safely explained that the study includes questions about violence and their formal consent to participate is requested. Consenting women participate in the study by completing a questionnaire and watching the video using an electronic tablet equipped with a headset. Women unable to read can use a data capture method known as a Color-coded Audio Computer Assisted Self Interview (C-ACASI) to give consent and complete the questionnaire, as described in our previous study (21). All women who complete the questionnaire at the first timepoint (questionnaire 1) will receive a phone call around 32 weeks gestation to invite them to complete a follow-up questionnaire (questionnaire 2).
The schedule of enrollment, intervention and data collection is displayed in Fig. 1.
The intervention
The intervention video is a further development from the flipchart intervention in our pilot study (21). The video lasts a total of 7 minutes and starts with a presentation of what violence is. To tell the story the video uses two fictive persons who tell about their experience of violence. In this way physical, emotional and sexual violence are presented using examples. The video informs women that they are not alone in this experience. The national helpline number and shelters are mentioned. It subsequently presents the same safety promoting activities (i.e. problem-focused coping strategies) as used in the pilot study including disclosing about the experienced violence (21). Women are told that it is not their fault that they are experiencing the violence. The list with suggested activities and number for the helpline is repeated. The video uses the drawings from the flipchart that have been animated as well as some new ones. The whole video is texted.
The control video also lasts 7 minutes and presents warning signs and symptoms women can experience during pregnancy that are associated with complications and risks for adverse pregnancy outcomes. This is a repetition of information provided during antenatal care sessions. Permission to use this video has been granted by MedicalAidFilms who produced it. This film was produced to raise community awareness to promote a healthy pregnancy (26). It too uses drawings and is texted in Nepali.
All women, irrespective of randomization arm receive a leaflet produced by the Nepal Health Authorities that promote a healthy pregnancy and informs women when to contact medical services.
The exposure
A culturally- and linguistically- relevant version of the Abuse Assessment Screen is used to measure the experience of DV (24, 25, 27). The original Abuse Assessment Screen for DV was developed in the USA (27), and has been widely used both clinically and in research (24, 28). The Nepalese-version of the screening tool has been validated by our team (publication forthcoming).
Outcomes
The primary outcome is the proportion of women who disclose someone outside the research team about their experience of DV. The outcome questions are common across both questionnaires to maximize comparison of data between the two time points.
Secondary outcomes are:
- Anxiety and depression: The ten-item Hopkins Symptom Checklist (HSCL-10) is used to measure symptoms of anxiety and depression. HSCL-10 demonstrates good sensitivity and specificity for detecting mental distress and is widely used, also in the Asian region (29, 30). Out of 10 items, 6 indicate symptoms of depression and four symptoms of anxiety. For each item the answering options range from not at all (1) to extremely (4). The rationale for measuring distress is twofold. Firstly, we presume that women hearing that violence is not their fault, that they are not alone, that it is not acceptable and that there are possible actions to be taken may become less distressed. Secondly there is the expectation that women who disclose someone about the violence will experience support and thus less distress.
- Coping strategies: The use of coping strategies will be measured using the Ways of Coping Checklist (WCCL) selecting items from the subscales measuring seeking social support (7 items), detachment (6 items) and self-blame (3 items) (31, 32). Assessing coping strategies using similar questions has been done in Nepal in several studies (33, 34). The response format is on a 4-point Likert scale, where 0=does not apply and /or not used, 3= used a great deal. We include these scales as we hope the information in the video will reduce the use of self-blame and detachment strategies and increase the use of support seeking strategies.
- Safety measures: The use of specific safety measures in relation to living with DV is measured using the list developed by McFarlane et al (2002) which we modified to the Nepali setting in our pilot study (21, 25).
- Attitudes towards abuse: Attitudes towards acceptance of abuse will be measured using 5 questions used by the WHO (35) and national health survey in Nepal (36). In addition will we include 6 items from the Social Acceptance of Wife Abuse Scale (SAWAS) which cover other situations and actions known to be found an acceptable reason for violence (37). The video tells women that they are not to blame for the abuse they receive. Thus, no actions on a woman’s part justify the use of violence.
Background and associated variables
We collect data on age, education, and income for the woman and her partner, if she has one. We also collect data on ethnicity, if they are living rurally or in a city and if they are living in a joint family or not. We also collect data on food insecurity using Household Food Insecurity Access Scale previously used in Nepal (38, 39). Food insecure pregnant women are more likely to suffer from violence from her household members (40). Physical activity (in relation to nutritional status) will be measured by modifying questions from the Global Physical Activity Questionnaire (41). Suffering from violence may have led to unintended pregnancy (42) so we collect data on pregnancy intendedness using London Measure of Unplanned Pregnancy Scale (43). These background and associated variables can potentially influence both the exposure and the outcomes.
Sample-size
In this study we will include 2000 pregnant women. The sample-size calculation is based on the findings from the national Demographic and Health Survey in Nepal 2016 and the validation study of the Nepal Abuse Assessment Screen (N-AAS) done in preparation for this study (publication forthcoming). The most recent Nepal survey at the time of planning the study reported that, among women who had experienced physical and sexual violence, 34% had talked to someone about it (36). The validation study showed that 12% of the pregnant women reported the experience of DV (publication forthcoming). To detect an increase of 20 percentage points from 35 to 55 % in the proportion of women who talk to someone about their experience of DV, the recruitment of 240 women experiencing DV will give at least 80 % power with a significance level of 5 % and after a presumed 15 % attrition rate. With an expected prevalence of 12% this requires the recruitment of 2000 women.
Randomization
The randomization on a 1:1 basis will be in blocks of 16. The randomization is computer generated, by recruitment site. The randomization program is installed by an external programmer. Upon consenting to participate in the study, women will be allocated into the intervention or control group irrespective of their DV status.
Blinding
Healthcare staff caring for the women are blinded to the allocation as long as the woman does not discuss the content of the video with them. The statistician who will perform the analyses for the outcome paper is blinded to the allocation. Women are blinded to the allocation as the recruitment room has been organized in such a way that women do not view other screens than their own. Recruitment staff, however, are not blinded to the allocation of the women and participants may consult them about their own situations and require referrals.
Process evaluation
The process evaluation will involve various participants, including 20-25 women who took part in the RCT, approximately 24 healthcare providers through Focus Group Discussion (FGD) and health managers from the gynecology and obstetrics departments. Additionally, interviews will be conducted with research assistants and doctoral researchers who are involved in the study following the guidelines of the Medical Research Council (MRC) framework (44) which includes considerations such as adaptations, reach, fidelity, and dosage.
The main outcome of the process evaluation is an assessment of the acceptability of the video intervention, specifically investigating the various factors that influence acceptability, such as the time required to explain and watch the video and safety considerations for the women involved. Additionally, the process evaluation will explore the factors that impact women’s disclosure of DV. This includes exploring their agency in seeking help, the relevance of the intervention content in facilitating effective assistance, and the different stages they experience during the process of seeking support for DV.
Methods: data collection, management, and analysis
Data collection
Participants who provide consent will complete a questionnaire using an electronic tablet. The questionnaire design has been adapted to accommodate and support women who have reading difficulties. The C-ACASI method enables women to listen to the questions and provide color-coded answers (45, 46). For women who do not require this assistance, they have the option to read and answer the questionnaire directly without listening to the questions, which can expedite the process for them. Women complete one questionnaire at baseline (Q1) and one approximately 10-26 weeks later (Q2). For completing the follow-up questionnaire, women will receive a phone call from the research assistant to confirm their previous participation in the study and extend an invitation for a second time during their regular antenatal visit. Additionally, research assistants will gather data from the patients’ obstetric notes, including both the antenatal card and hospital register.
Analysis
A detailed statistical analysis plan (SAP) will be developed.
The main outcome is measured by comparing the proportion of women in the intervention and control groups who have told someone about the violence they have experienced. The comparison will be presented as risk ratios with 95 % confidence intervals. In addition, we will collect data on whom the woman has told about the violence and will present this as descriptive findings. A detailed description of the analysis strategy for the secondary variables will be provided in the SAP. Briefly, binary outcomes will be compared between intervention arms using risk ratios, and continuous secondary outcomes will be compared using linear mixed models to take into account the baseline values of each outcome variable. Non-parametric analysis or alternative regression models will be considered where the assumptions for the linear regression models are not met.
Process evaluation
The observations of the RCT and individual in-depth interviews with participating pregnant women will be conducted at different stages during the implementation of the RCT and upon its conclusion. The FGDs with healthcare providers and health managers will take place at the end of the RCT, while interviews with research assistants and doctoral students will occur after the analysis of RCT outcomes. The interviews will be recorded, transcribed, and analyzed using inductive coding to derive themes, theories, and concepts from the data, thereby revealing underlying structures of experiences or processes (47).
Data management and analysis will follow a convergent parallel mixed method design. Qualitative and quantitative data will be analyzed separately and then merged and integrated at the interpretation level. Data triangulation, including the use of a triangulation matrix and discussions with co-researchers, will be employed to ensure a comprehensive analysis.
Adverse events and data monitoring
The study will be supervised by an internal Data Safety and Monitoring Committee (DSMC). A safety protocol has been set up and adverse events will be reported to the committee as they occur. The safety protocol includes immediate referral opportunities for women depending on their individual needs. The characteristics of the trial are such that an external data monitoring and safety committee was considered unnecessary. These characteristics include a low risk of harm to the participants due to the trial and negligible credibility/integrity concerns about the potential impact (48). No interim data analyses are planned. The DMSC will monitor the quality of the data.
Study management
The study is managed by the Principal Investigator (PI) and co-PI in Norway and a local PI at each of the participating hospitals. This team is responsible for ensuring the study is conducted according to the protocol and ethical approval, that women’s privacy is protected, and that women in need of referral or other support receive this. Data-management is regulated by collaborative standard contracts.
Dissemination
A dissemination event will be organized in Nepal where findings of study will be presented. This event will be open to all stakeholders. The video will be handed over to the National Health Education, Information and Communication Centre at the Department of Health Service, Nepal. We will explore rural uptake of this intervention during the dissemination event. All papers will be submitted for publication in international open access scientific journals.