Eligibility criteria

Studies meeting the following criteria were included: Primary research (original articles excluding systematic reviews), targeted at IPV survivors, describing interventions focused on improving access to resources and mental health outcomes for IPV survivors. The interventions had to use a social support or network-oriented approach. There were no restrictions on gender, but most of the studies identified focused on female survivors of violence (See Table 1). We defined ‘IPV as physical, sexual and psychological abuse directed against a person, by a current or ex-partner’ [10].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Summary of studies selected.

https://doi.org/10.1371/journal.pone.0235177.t001

Studies had to address the following outcomes: intimate partner violence, social support, mental health outcomes and quality of life. Other outcomes that were also included were those associated with access to resources, utilisation of health services, and safety-promoting behaviours, if they were assessed in addition to the outcomes mentioned earlier. No restrictions were placed on study design or language, to allow for inclusion of all relevant studies.

Information sources

Between May and July 2017, we conducted a search across 5 databases: Medline, Embase, Web of Science, Cochrane and PROQUEST, for studies published between 1980 and 2017. We decided to include studies from the 1980’s because some of the pioneering publications on the use of advocacy and social support, for example, Sullivan et al’s work were published in the late 80’s and early 1990’s and we wanted our review to include some of these publications. Even though the review eventually included only primary studies, we included studies from COCHRANE to allow us to identify additional articles. We did not conduct a separate search for grey literature, as the PROQUEST database also included scholarly journals, newspapers, reports, working papers, and datasets along with e-books. Retrieved references were imported to Endnote and Mendeley and were then transferred to a systematic review software called Co-evidence [11]. In January 2019, another search was done to update and ensure new articles or information could be included in the review. Table 1 provides an overview and summary of the studies selected, as well as the evidence ranking of the studies.

Search strategy

The search strategy was developed in collaboration with a librarian, as well as a review of other existing systematic reviews on IPV or social support interventions. Search terms combined MeSH terms, and specific terms related to IPV and were adapted to each of the databases searched. This is presented in Table 2.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Search terms and articles retrieved from each database.

https://doi.org/10.1371/journal.pone.0235177.t002

Study selection

Inclusion of retrieved studies and their eligibility were independently assessed by two reviewers, EO and SH, in a two-step process. First, the authors independently screened all study titles and abstracts using Co-evidence (the systematic review software), which notified each author of conflicts. When a conflict was identified, articles were again independently reviewed, and discordance was resolved through discussion, using the systematic review protocol as a guide. The same process was also used for the full text-screening phase of the study. While this process lengthened the screening process, it allowed for transparency and made it possible for both reviewers to continually reference the study protocol and ensure that the study objectives were adhered to, through the review process.

Data extraction

A standardized data collection form was developed by EO and SH, adapted from the Cochrane data collection grid. EO extracted all the data from the studies, SH and RB reviewed the data and it was agreed that OD would provide input if there was any disagreement about the data extracted.

Risk of bias

The quality and risk of bias in the studies were independently assessed by EO and SH, using the appropriate quality assessment tool. As the studies selected included quantitative and qualitative studies, there was an agreement to assess quantitative and qualitative studies separately. Quantitative studies were assessed using the Quality Assessment Tool for quantitative studies developed by the Effective Public Health Practice Project, see Table 3 for an overview of the components of this tool [12]. This tool had been used in another systematic review focused on interventions [13]. Qualitative studies were assessed, using the Critical Appraisal Skills Programme (CASP) Qualitative Research Checklist [14], the main components focused on assessing the methodological limitations, coherence, adequacy of data and relevance of research. See Table 4 for an overview.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Ranking criteria used for quantitative studies: Effective public health practice project: Quality assessment tool for quantitative studies. [12].

https://doi.org/10.1371/journal.pone.0235177.t003

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. CASP qualitative studies checklist.

https://doi.org/10.1371/journal.pone.0235177.t004

Information about studies selected

The initial search across the different databases retrieved 3712 articles, of which 3364 articles were irrelevant based on the screening criteria. 337 articles were assessed at the full text screening stage, and 27 articles selected to be part of the systematic review, the overview is presented in Fig 1

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. PRISMA flow diagram for the systematic review of IPV interventions focused on improving social support and mental health outcomes among IPV survivors.

From: Moher D, Liberati A, Tetzlaff J, Altman DG, The PRISMA Group (2009). Preferred Reporting Items for Systematic Reviews and Meta-Analyses: The PRISMA Statement. PLoS Med 6(7): e1000097. doi: 10.1371/journal.pmed1000097 For more information, visit www.prisma-statement.org.

https://doi.org/10.1371/journal.pone.0235177.g001

Results/Key findings from the systematic review

The interventions were classified based on the methodology or type of social support provided to the survivors of violence. Most of the studies identified involved the use of an ‘advocate/ case manager’ or ‘interventionist’ (which referred to a nurse, psychologist or volunteer trained to administer the IPV intervention). The advocate was often responsible for offering the survivor information on resources and helping them identify safety strategies. The interventions usually consisted of weekly sessions or phone calls for a certain period of time. These interventions were mostly in the United States and from other countries like China, Canada, Denmark, Netherlands, Uganda and the United Kingdom. Other interventions involved the use of advocacy with an added psychotherapy component, and interventions that focused on community education, as well as empowerment of the IPV survivors. One of such community focused interventions used an empowerment model and encouraged survivors of violence to take photos of their safety strategies. These photos were used to educate the community about the consequences of intimate partner violence and advocate for community support to prevent intimate partner violence and encourage access to services. In our paper, the term ‘community focused’ included interventions targeted at the community which used participatory and non-participatory methods in the design and implementation of the programmes. The interventions identified in this systematic review had different target groups, pregnant women, survivors of violence resident in shelters, community members and IPV survivors, substance abusing women, and women with small children.

Types of social support interventions for intimate partner violence survivor

Advocacy/ Case management interventions

These interventions involved the use of community-based advocacy interventions focused on individuals that were survivors of violence, these interventions were focused on assisting the survivors identify and access resources, supportive relationships and cope with the effects of intimate partner violence. Fifteen of the studies reviewed (11 RCTs, 2 pre-post evaluation, 1 retrospective study, 1 quasi-experimental study with randomization) described experiences with social support interventions that provided some sort of advocacy service in combination with community support for survivors of violence, on an individual level [15–29].

Advocacy interventions may include ‘helping abused women to access services, guiding them through the process of safety planning, and improving abused women’s physical or psychological health’ [30]. For the review, interventions grouped under this category included mentor-mother interventions (these interventions involved the training of IPV survivors who were mothers as counsellors and mentors, for other IPV survivors), and use of home-based or in-clinic advocates. Most of the studies reported a decrease in depression, fear, post-traumatic stress disorder, and increased access to social support for the IPV survivors included in the study.

In Tiwari et al’s study, where an advocacy intervention was compared to the usual community services, the reduction in depression and other mental outcomes, was not significant but the reduction in ‘partner aggression’ and increase in access to social support in the intervention arm was significant [15]. Two of the studies, an in-clinic advocacy intervention by Coker et al [23] and a home-based advocate intervention by Sharps et al [20] resulted in a significant reduction in the experience of intimate partner violence by the survivors (decrease in experience of IPV in the intervention arm compared to the control group). The two mentor mothers’ studies included in this review, showed an increase in uptake of support services and mental health services. Prosman et al’s study [18] specifically showed evidence that the mentor mother intervention led to a decrease of in experience of IPV (decreased Composite Abuse Scale (CAS) mean score by 37.7 (SD 25.7) after 16 weeks), as well as in depression scores. This study had a component that focused on uptake of therapy, which may have influenced the outcomes. Four of these studies compared ‘face to face’ case management/ advocacy services to remote modes of care and assessed the impact on social support and IPV. Gilbert et al’s study [24] compared online and case manager implemented screening, assessment, and referral to treatment intervention for IPV survivors who were substance abusing, the intervention was guided by social cognitive theory, and focused on short screening, an intervention and referral to treatment (SBIRT) model. There were no significant differences between both groups in terms of impact of the interventions, the study found both groups has an increase in access to social support, IPV self-efficacy (ability to protect themselves from IPV) and abstinence from substance use, irrespective of the type of intervention they received. McFarlane et al [26] assessed the differences between nurse case management and a referral card on reduction of violence and use of community resources among IPV survivors, and found no differences in outcome between both groups, but found compared to baseline, participants who received either intervention (nurse case management or referral card) had a significant reduction in experiences of violence (threats of abuse, assaults, risks of homicide and work harassment) between baseline and 24 months post-intervention. There were no significant differences in outcome for participants who were in the referral card or case management intervention arm. Other outcomes like improved safety behaviors and a reduction in the utilization of community resources were also found across both groups. Stevens et al’s [27] study focused on using telephone based support/referral services for IPV survivors compared to enhanced usual care (, the intervention was based on a social support and empowerment model. The study found no significant difference in outcomes between the intervention arm (telephone-based arm) and the control arm (enhanced usual care- community services provided by the community center including health, social, educational, and recreational services). Research participants reported a decrease in experiences of IPV across both groups, associated with ‘higher levels of social support’ at baseline and at 3 months post-intervention. However, the reduced levels of violence did not influence the capacity to obtain or utilize community resources among the research participants. Constantino et al’s [29] study compared an advocacy based intervention across different methods (online and face to face) and found the intervention reduced depression, anxiety and increased personal and social support among the online group compared to the control group. The intervention included a module that addressed interpersonal relationships, thoughts and emotions as well as access to referral services like legal aid. Another study by Constantino [28] involved a nurse led intervention focused on providing information on resources and services for IPV survivors living in a domestic violence shelter. The intervention was compared to usual care in the shelter. The intervention group had reduced psychological distress, increased levels of social support and reduced reporting of health care issues. Most of the studies we found in this category showed moderate levels of quality of evidence.

Advocacy/Case management interventions with a psychotherapy component

3 of the studies (3 RCTs) [31–33] were focused on interventions that included specific types of psychotherapy, sometimes delivered remotely or through individual or group sessions. Zlotnick et al [31] described the use of interpersonal psychotherapy among pregnant women focused at improving social support among the survivors of violence during individual psychotherapy sessions. Though there was a moderate change in depression and PTSD scores (reduction) between the control and intervention groups at post-intake (5–6 weeks), this difference was not sustained at the post-partum period. Hansen et al [33] describes the use of psychotherapy using either the ‘Trauma Recovery Group’ (TRG) method developed by ‘a private Danish organization called ‘‘The Mothers’ Aid”‘ or regular trauma therapy for individual or groups of women who were survivors of IPV. The study reported significant changes in PTSD, depression and anxiety symptoms and increased levels of social support (high effect sizes); however, our assessment with the EPHPP grading revealed that the study design was weak. Miller et al’s [32] study shows the effect of a ‘mom empowerment programme’ focused on improving mental health outcomes and ability to access resources among IPV survivors participating in the programme, with resulting improvement in PTSD, depression and anxiety symptoms.

Community-focused/ network social support interventions

These group of studies, distinct from the ones described above focused on community education and change, so the focus of the studies was not just the individual survivor of violence, but the community as a whole. 9 (3 RCTs, 3 pre-post evaluations, 3 qualitative research) of the studies we reviewed consisted of interventions described as being community-based [34–42]. The definitions of community-focused interventions used for classifying the studies followed the typology by McLeroy et al [43], which refers to interventions where:

• The setting of the intervention is the community
• The target population of the intervention is the community
• The intervention uses community members as a resource
• The community serves as an agent for the intervention (i.e. interventions working with already existing structures within the community)

We have focused on interventions in this category where the focus of the intervention is the community. The interventions described include community participatory research, like those described by Ragavan et al’s systematic review on community participatory research on domestic violence [44], as well as interventions that are ‘community placed’, where the community is a target of the intervention, and might not have been involved in the design of the intervention, in a participatory way.

All the interventions were focused on IPV reduction and improving social support and mental health outcomes for survivors of violence. Interventions like SASA [34, 39], used community members as a resource for the intervention. In the SASA intervention, community activists in the intervention sites were trained on GBV prevention, power inequalities and gender norms. After training, they carried out advocacy activities, engaging different stakeholders and members of their social networks to address harmful social norms around GBV. At the end of the intervention, there were reported lower rates of IPV among the intervention community. Other interventions like the ‘Framing Safety project’ [35], which focused on promoting agency and self-empowerment among survivors of violence, found that by providing means through which survivors of violence could tell their own stories and take ownership of this process, there was a resulting feeling of empowerment among the women. Other interventions used group therapy sessions that were community-based and culturally tailored to the specific target population. Wuest et al [41] described a collaborative partnership with different stakeholders (academic, NGOs and community members) to develop a comprehensive intervention to IPV, ‘Intervention for Health Enhancement After Leaving (iHEAL), a primary health care intervention for women recently separated from violent/abusive partners’. The post evaluation revealed significant reduction in depression and PTSD from baseline to 6 months post-intervention, these improvements in mental health outcomes, were present at 12 months post-intervention. Other outcomes, like social support, showed some initial improvement from baseline to 6 months post-intervention but these changes were not sustained till 12 months post-intervention.

Community focused/ network interventions with a psychotherapy component

Three of the nine studies (1 RCT and 2 pre-post study) by Kelly et al [36], McWhirter et al [37], and Nicolaidis et al [38] described group therapy interventions that were designed in collaboration with the target population in a participatory way. These studies reported significant reductions in severity of mental health conditions like depression and PTSD, as well as an increase in social support and self-efficacy for the women who were involved in the study.

Summary of key findings and recommendations

• Most of the interventions that used advocacy with strong community linkages and a focus on community networks showed significant effects on mental health outcomes and access to social support, we assume a reason for this could be that because these interventions were rooted in the community, there were more sources of support that allowed the survivors of violence to develop better coping strategies, for example in the SASA study that included a strong community engagement component, community responses to cases of IPV were supportive of the survivor, and this had an effect on incidence of IPV. Future research and interventions on IPV should focus on ensuring stronger community linkages and outreach programmes to enhance the impact of the interventions on IPV survivors.
• This review found that when remote modes of intervention delivery were compared to ‘in person’ delivery of an intervention, there were no significant differences in outcome. This finding is of specific importance to hard-to-reach and vulnerable populations whom might be unwilling to access care at hospitals and registered clinics. More research focused on the use of remote support interventions among vulnerable populations (specifically IPV survivors), should be encouraged.
• There was a lot of heterogeneity in outcome measurements, especially measures of social support, drawing attention to the need for research and discussions around standardization and synthesis of evidence-based research on social support and IPV.
• In some of the studies, the ‘dosage of the intervention’, as well as some participant characteristics like age or ethnicity are often cited as potential moderators of some of the outcomes, more research on IPV intervention should examine the time dependent nature of interventions and their effect on outcomes similar to what was done by Bybee et al[16].

Limitations

Although there were no language restrictions included in our search strategy, most of the studies retrieved and subsequently reviewed were in English, which could have influenced some of our conclusions.