A systematic review of intimate partner violence interventions focused on improving social support and/ mental health outcomes of survivors

Background Intimate partner violence (IPV) is a key public health issue, with a myriad of physical, sexual and emotional consequences for the survivors of violence. Social support has been found to be an important factor in mitigating and moderating the consequences of IPV and improving health outcomes. This study’s objective was to identify and assess network oriented and support mediated IPV interventions, focused on improving mental health outcomes among IPV survivors. Methods A systematic scoping review of the literature was done adhering to PRISMA guidelines. The search covered a period of 1980 to 2017 with no language restrictions across the following databases, Medline, Embase, Web of Science, PROQUEST, and Cochrane. Studies were included if they were primary studies of IPV interventions targeted at survivors focused on improving access to social support, mental health outcomes and access to resources for survivors. Results 337 articles were subjected to full text screening, of which 27 articles met screening criteria. The review included both quantitative and qualitative articles. As the focus of the review was on social support, we identified interventions that were i) focused on individual IPV survivors and improving their access to resources and coping strategies, and ii) interventions focused on both individual IPV survivors as well as their communities and networks. We categorized social support interventions identified by the review as Survivor focused, advocate/case management interventions (15 studies), survivor focused, advocate/case management interventions with a psychotherapy component (3 studies), community-focused, social support interventions (6 studies), community-focused, social support interventions with a psychotherapy component (3 studies). Most of the studies, resulted in improvements in social support and/or mental health outcomes of survivors, with little evidence of their effect on IPV reduction or increase in healthcare utilization. Conclusion There is good evidence of the effect of IPV interventions focused on improving access to social support through the use of advocates with strong linkages with community based structures and networks, on better mental health outcomes of survivors, there is a need for more robust/ high quality research to assess in what contexts and for whom, these interventions work better compared to other forms of IPV interventions.


Introduction
The global prevalence of intimate partner violence (IPV) has been estimated at about 30% for women aged 15 and over [1]. We define IPV within this paper as 'any acts of physical violence, sexual violence, stalking and psychological aggression (including coercive tactics) by a current or former intimate partner' [2]. IPV affects men and women, and men or women can be perpetrators or survivors of violence. However, women are the most affected by IPV, and men tend to perpetrate violence more than women [3]. Survivors of violence are likely to first disclose experiences of intimate partner violence and expect informal support from a friend, family member, neighbour or other members of their social network, prior to seeking support from formal sources like health institutions and legal officers, however, the extent of disclosure differed with age, nature, ethnicity and gender [4].
IPV has been found to be associated with an increased risk of poor health, depressive symptoms, substance use, chronic disease, chronic mental illness and injury for both men and women [5]. Social support has been found to be an important factor for mediating, buffering and improving the outcomes of survivors of violence and improving mental health outcomes [6]. Conversely, social isolation and lack of social support have been found to be linked with poor health outcomes for survivors of violence. Liang et al [6] discussed the importance, perception of the abuse by the IPV survivor plays on their decision to ask for help and support. They mentioned how cultural factors including stigma and shame around disclosing IPV, perception of the incident as a personal problem and awareness of resources available, play a determining factor on types of resources accessed, especially for IPV survivors with a migrant background or of a low socioeconomic status. IPV survivors who perceive the abuse to be a personal problem were more likely to use placating and avoidant strategies before seeking external support [6].
In this study, we make use of Shumaker and Brownell's definition of social support, and define it as any provision of assistance, which may be financial or emotional, that is recognized by both the beneficiary and provider as advantageous to the beneficiary's welfare. ' [7]. IPV interventions that involve the use of social support, have the potential to improve the health seeking behaviour, access to resources and mental health outcomes of IPV survivors. Commonly cited types of social support interventions include but are not limited to the use of peer support, family support and the use of 'remote interventions like the use of internet or telephones as sources of social support from trained counsellors, as well as information about resources' [8]. Goodman and Smyth [9] discussed the importance of using a 'network oriented' approach to provision of domestic violence services that takes into account the value of informal support, from social network members of IPV survivors, as this would promote the well-being of the survivor and sustain some of the benefits of the intervention over time. Given the existing gap in evidence on the effect of different IPV interventions on social support and/ mental health outcomes of IPV survivors, this study aimed to address the evidence gap, by assessing the effects of these different IPV interventions, and network oriented approaches on improving access to social support and improved mental health outcomes for IPV survivors. This is of added benefit, as access to social support improves the mental health outcome of survivors of violence. More evidence of different types of social support interventions targeted at different groups of people, that are effective in addressing mental health outcomes of survivors, are needed.

Methods
The systematic review was developed according to the PRISMA (Preferred Reporting Items for Systematic reviews and Meta-analyses) guidelines. The methods used to screen the studies and define eligibility are described below:

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].
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 Coevidence [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. The presence of social support and resources were found to be significant mediators of 'quality of life' scores in the 'control' and 'intervention' groups. Social support had more influence on outcomes (as a mediating factor) than access to resources for the research participants (p = 0.003). However this effect was not sustained over 24 months. At this time, social support played less of an influential role in outcomes, and access to resources had a significant effect. Women in the intervention arm (who worked with advocates) had positive outcomes sustained up to 24 months (quality of life, social support, access to resources and decline in experiences of repeated abuse). of informing others about the realities of these experiences, using photography. The study reported an increased feeling of empowerment among women who participated in the project. The fact the intervention occurred within a support group context, seemed to make a difference for the participants, as this was a source of support and confidence. They were more willing to share their experiences of IPV with a public audience (using an exhibition) and be advocates for IPV prevention. There were no significant differences between groups across all outcomes, i.e.

Moderate
No added benefit of using a nurse case manager. Two years following treatment, both treatment groups of women reported significantly (p < .001) reduced experiences of abuse threats, assaults, harassment at work and risks of homicide but there were no significant differences between groups. Compared to baseline, both groups of women adopted significantly (p < .001) more safety behaviours by 24 months (however, community resource use declined significantly (p < .001) for both groups. There were no significant differences between groups. The programme was effective in improving the number of referrals among the women who used the advocacy services, and was reported to be effective 'in offering social support, education and assistance with needed community resources'.

Weak
(Continued ) Herman (1997), the intervention program in this study followed a three-phase stepwise treatment approach, 1) Stabilisation program, 2) Treatment program, 3: the follow-up program respectively.
Stabilization program phase: Participants who completed this phase had significant reductions in PTSD, depressive symptoms and anxiety symptoms (from baseline to postintervention). They also had a significant increase in levels of perceived social support from baseline to completion. Participants who completed the treatment program experienced significant reductions in PTSD depressive symptoms and anxiety symptoms (from baseline to completion of the treatment program). There was no significant difference in levels of perceived social support.

Weak
18. Differential therapeutic outcomes of community-based group interventions for women and children exposed to intimate partner violence.

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.

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.

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

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 or 'date rape' or battering or victimization or rape or ' intimate partner violence' or 'sex offense � ' or 'sexual violence' or 'sexual assault � ' or 'sexual violation' or ((partner or spous � or domestic or dating or couple or marital or wife or wive � ) near/2 (abuse or aggression or violence or assault or maltreatment)) or rape or 'battered wom?n' or (gender � near/2 violence) or (violence near/2 wom?n) 169 AND (Link � near/2 (care or healthcare or service � or therap � or treatment � )):ti,ab or (access � near/2 (care or healthcare or service � or therap � or treatment � )):ti,ab or (utiliz � near/2 (care or healthcare or service � or therap � or treatment � )): ti,ab or (seek � near/2 (care or healthcare or service � or therap � or treatment � )):ti,ab or 'health intervention' or 'pilot study' or 'public health' AND [mh^"Social Support"] or [mh "community networks"] or "social network" or "social support" or "community support" or ("support system � " or "social support" or "community support" or "social network � " or "community network � " or "community involvement" or "family support" or "family involvement" or "family network" or "parental support" or "parental involvement" or "parental support" or assistance or encouragement or coping or "better cope" or "peer involvement" or "peer support" or "peer network � " or friends):ti,ab (Continued ) AND (("support system � " OR "social support" OR "community support" OR "social network � " OR "community network � " OR "community involvement" OR "family support" OR "family involvement" OR "family network" OR "parental support" OR "parental involvement" OR "parental support" OR assistance OR encouragement OR coping OR "better cope" OR "peer involvement" OR "peer support" OR "peer network � " OR friends):AB,TI) https://doi.org/10.1371/journal.pone.0235177.t002 Was the recruitment strategy appropriate to the aims of the research?
Was the data collected in a way that addressed the research issue?
Has the relationship between researcher and participants been adequately considered?

Have ethical issues been taken into consideration?
Was the data analysis sufficiently rigorous?
Is there a clear statement of findings?

Section C: Will the results help locally
How valuable is the research?
https://doi.org/10.1371/journal.pone.0235177.t004 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 nonparticipatory 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
Survivor-focused social support interventions. The interventions described below were all focused on providing social support and improving mental health outcomes for the survivors of violence, all of them involved the use of advocacy/case management approaches, through remote or 'face to face' methods. We also identified advocacy interventions with a strong therapeutic component, which we have discussed separately.

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][16][17][18][19][20][21][22][23][24][25][26][27][28][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
(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][35][36][37][38][39][40][41][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.

Discussion
The focus of this systematic review was to assess the existing evidence available on IPV interventions focused on improving social support and/or mental health outcomes. To ensure that we included all relevant studies, we included both quantitative and qualitative articles. 27 articles were included in the systematic review out of 337 full text articles assessed. The following interventions were identified via the review: Survivor focused interventions (18 studies: 15 of these studies were focused on advocacy/case management services; 3 of these on advocacy/ case management services with a psychotherapy component), community-based social support interventions (9 studies:4 out of these were community coordinated interventions with a psychotherapy component). The heterogeneity of the studies made it difficult to conduct a meta-analysis because of the variability in outcome measures, study design and processes and duration of interventions implemented. Survivor focused advocacy/case management IPV interventions made up most of the interventions identified (18 out of 27). The studies showed good to moderate evidence of the positive impact of these interventions on mental health outcomes and also access to social support for the IPV survivors included in the study, and in a few studies, a reduction in partner aggression or experience of IPV (IPV scores) [15][16][17][18][19][20][21][22][23]. In one study, by De Prince et al [42], where a community-based advocacy intervention was compared to an advocacy intervention that was focused on referral, both groups showed improvement in mental health outcomes, but the community-based advocacy intervention group (outreach) had slightly better mental health outcomes. A specific approach of the intervention was that it was community-led/ coordinated, the community based organisation reached out directly to the survivors of violence based on information from the systems based advocate, hence removing the need for survivors to seek out services themselves based on the referrals received from the system based advocate. This study might have important lessons for future advocacy interventions, as just provision of referrals might not ensure uptake of services, and a community coordinated follow up of IPV survivors might be more effective in ensuring uptake. However, it must be noted that only few of the advocate-based studies and 1 of the community-focused interventions reported an impact on IPV, with good level of evidence [15,[20][21][22][23]34], similar to what has been found in other reviews of advocate-based interventions on intimate partner violence [45]. Tiwari et al's study, which focused on the use of an empowerment, social support and advocacy-focused telephone intervention, found improved mental health outcomes among the intervention group. In comparison, Cripe et al's [46] study also compared the effect of an empowerment-based intervention in comparison to usual care among abused pregnant women and found higher scores of improved safety behaviours among the intervention group compared to the control group but 'no statistically significant difference in health-related quality of life, adoption of safety behaviours, and use of community resources between women in the intervention and control groups'. These differences we attribute to the study design, context and characteristics of the study participant. Goodman et al has described the importance of integrating a 'social network' approach into IPV interventions, and linking interventions with social networks of IPV survivors to ensure sustained access to social support for the survivors [9,47]. Many of the advocacy/case management interventions described above have created these linkages by assisting IPV survivors identify sources of support within their existing networks and also engage in forming new social relationships [16,18,48]. However, more IPV interventions should integrate this approach in a coordinated systemic manner, as engaging with social network members of the IPV survivors ensures sustainability of the programme's effects over time [9].
Several of the studies focused on psychotherapy interventions, which were individual, or group based. We classified these interventions separately as these interventions combined community-based advocacy with a therapeutic component, as opposed to advocacy/case management alone or community focused interventions. These interventions either used interpersonal therapy [31], traumatic treatment therapy [33], empowerment based group therapy [32], and a multicomponent intervention that combined therapeutic education sessions with information on resources and legal help remotely or 'face to face' [29]. All the interventions showed some impact on mental health outcomes and social support, with a weaker level of evidence of an impact on IPV. Although Zlotnick et al's study [31] on a therapeutic intervention for pregnant IPV survivors, described an improvement of mental health outcomes (moderate effect on PTSD and depression), this finding was not sustained in the postpartum period, drawing attention to the need to assess the efficacy of interventions in this particular group, taking into account time dependent factors and participant attributes. A review done by Trabold et al [49], found that clinically focused interventions and group-based cognitive or cognitive behavioural interventions had a significant effect on depression and PTSD, as well as the uses of Interpersonal therapy (time dependent). However, as our review focused on therapies focused on improving social support and mental health outcomes, we included fewer studies. Although we found a similar trend as described by Trabold et al, among community-based interventions (including those that were psychotherapy focused), we could not assign the effect specifically to the type of psychotherapy method, but rather to the length, associated support services and context of the intervention. Sullivan et al [50] discussed the positive effect of trauma informed practice on mental health outcomes of IPV survivors in Shelters, showing evidence of the importance of IPV interventions to include a comprehensive 'therapeutic or mental health component'. They also discussed the six components of what 'trauma informed practice' which includes: (a) reflecting and understanding of trauma and its many impacts on health and behaviour, (b) addressing both physical and psychological safety concerns, (c) using a culturally informed strengths-based approach, (d) helping to illuminate the nature and impact of trauma on survivors' everyday experience, and (e) providing opportunities for clients to regain control over their lives'. These components were useful for advocacy/case management interventions for IPV survivors, to ensure a focus on improving mental health outcomes, intersectional collaboration between stakeholders, and that the intervention is survivor-centred and addresses cultural factors.
Interventions that compared remote and 'face to face' methods of support and advocacy mostly resulted in a reduction in IPV victimization and increased access to social support. In cases where different modes of intervention delivery were tested, for example a comparison between remotely delivered interventions (telephone or online) and 'face to face' interventions, no difference was noted between both modes of intervention. Krasnoff and Moscati's study [51] discussed a multi-component referral, support and case management intervention that reported similar reduction in perceived IPV victimization as seen in studies included in our review. There were some differences in the telephone support interventions included, Stevens et al's study [27] reported no difference in mental health outcomes compared to Tiwari et al's study [15] which found an improvement in mental health outcomes among the intervention group. We postulate differences in outcome could be attributable to the fact that Tiwari's intervention was more advocacy, empowerment and support focused than the intervention described in Stevens et al study, which was more information and referral focused.

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.

Conclusions
This systematic review presented the findings from IPV interventions focused on social support and mental health outcomes for IPV survivors. Advocacy/case management interventions that had strong linkages with communities, and were community focused seemed to have significant effects on mental health outcomes and access to resources for IPV survivors. However, all IPV survivors are not the same, and culture, socioeconomic background and the perception of abuse by the IPV survivor, have a mediating effect on their decision to access social support and utilize referral services. 'An intersectional trauma informed practice' [50] [52] that addresses psychological and physical effects of IPV, is culturally appropriate and is empowering for the survivor, in addition to a 'social network oriented approach' might provide a way to ensure that IPV interventions are responsive to the needs of the IPV survivor [47]. This will ensure the interventions are targeted at ensuring survivors are able to access social support from their existing networks or new social relationships, and might also promote community education about IPV and promote community support for IPV prevention and mitigation.
Future studies on IPV interventions should assess how these approaches impact the incidence of IPV, social and mental health outcomes across different populations' of IPV survivors.