Intimate partner violence and post-migration stressors reported by refugee women accessing settlement services

Abstract

Intimate partner violence (IPV) is highly prevalent globally, with increased risk for women in situations of conflict, post conflict and resettlement. The Safety and Health after Arrival (SAHAR) study tested IPV screening with women accessing settlement services in New South Wales, Australia, using the validated ACTS tool, along with brief response involving risk assessment, safety planning and referral. A three month follow-up telephone survey was administered to women who had attended four participating sites which delivered the intervention. The survey explored the nature of any IPV experienced, factors associated with disclosure, and responses provided to those who identified IPV. Data is reported on 316 women of whom 48 (15%) identified current IPV. For 45 women who responded to Composite Abuse Scale items, the most common forms of abuse were forced isolation from family/friends (56% 25/45), blame for abusive behaviour (53% 24/45), “put downs” (44% 20/45) and physical violence 38% (17/45). Psychological distress and post-migration stressors were significantly higher for women who disclosed IPV compared to those who did not. Length of residency in Australia and whether the screening occurred during the first or subsequent service visits, were not associated with the likelihood of disclosing IPV. The majority of women who disclosed reported the caseworker’s response to be helpful and involved risk assessment, safety planning and referral. Screening and response to disclosure in settlement services provide opportunities to address abuse experienced by this group of women who are less likely to report experiences of abuse or use mainstream services.

Introduction

Intimate partner violence (IPV) is defined by the World Health Organization as a ‘pattern of behaviour by a current or former intimate partner that causes physical, sexual and psychological harm, including acts of physical aggression, sexual coercion, psychological abuse and controlling behaviours’ [1]. Experienced by 26% of ever partnered women globally [2], IPV leads to severe physical and mental health impacts, as well as substantial social and economic costs [3,4,5].

UNHCR and a number of countries work to resettle refugees who cannot return to their home country because of continued conflict, wars and persecution. Resettlement involves refugees moving to a third country, which has agreed to admit them as a durable solution to their protection. Women in resettlement contexts are at increased risk of IPV due to heightened isolation, adjustment and acculturation stress, non-citizen status, language barriers, economic insecurity, and limited social support [6, 7]. IPV prevalence estimates for refugee and migrant women are limited and vary across jurisdictions [8]. In Australia it is estimated that one third of refugee and migrant women have experienced domestic and family violence [9] with evidence that rates of controlling behaviour and intimate partner psychological violence are higher for women who arrived as refugees compared to Australia-born women [10]. Immigrant women are more likely than those born in the USA to be killed by an intimate-partner [11] and recent Australian data indicates that 27% of those killed by their partner were born in another country [12]. At the same time, former refugees are less likely to report experiences of abuse to police and are more likely to remain with abusive partners than locally-born women [13,14,15,16]. This situation reflects socially determined structural inequities which include unequal access for refugee women to the means to address IPV [17], putting women at risk of violence by those who would take advantage of their precarity.

At the individual and community level, separation from families, pre-arrival trauma, and limited support networks exacerbate vulnerability for this group of women [18, 19] who are also less likely to use mainstream health services [18, 20, 21]. Lack of knowledge of local laws and systems, visa precarity and language barriers, create additional impediments to help-seeking [22, 23]. Nevertheless, it is important to note that former refugees are typically resourceful during settlement [24], including those experiencing IPV, who exhibit choice and agency, drawing on individual, family and community strengths [25]. Settlement services in countries of resettlement provide information and support for those who have arrived through forced migration [26].

Women who re-settle in a new country after forced migration face additional challenges. While pre-migration stressors are well understood to be cause for poor mental health [27], stressors in the post-migration period are increasingly recognised as significant to psychological functioning [28,29,30]. Depression and anxiety are common experiences for former refugees [31]. There is value in understanding how these factors interact with IPV, given their potential to exacerbate impacts of forced migration and challenges of resettlement.

Universal screening or inquiry for IPV which involves asking all women attending designated services a small number of standardised questions about recent experiences of IPV is recommended for priority populations [32]. Directly asking about experience of violence increases disclosure and creates opportunity for supportive intervention [33, 34]. IPV screening using validated tools has been implemented in diverse health settings including: ante-natal clinics, primary health care, emergency departments, well baby clinics, substance treatment programs and mental health services [35, 36]. Settlement services potentially offer safe opportunities to identify IPV among former refugees, however only one prior study has been identified which explored use of screening in settlement services [37].

The Australian Government funds settlement programs to assist refugees in their resettlement and help them access mainstream services and programs available to other Australian residents across a range of integration domains [38]. The Humanitarian Settlement Program provides case management for the first 18 months post-arrival. The follow-up program, Settlement Engagement and Transition Service (SETS) supports former refugees and some eligible marginalised migrants, from 18 months to five years post-arrival [39].

This study introduced and evaluated a culturally tailored IPV screening and response strategy within a major settlement program with newly arrived refugee women. The intervention to identify and respond to IPV was introduced in four settlement services in NSW Australia and comprised instituting the ACTS tool [40], providing a wallet-sized IPV information card translated into community languages, offering referral to an on-site dedicated IPV worker trained and supported to deliver risk assessment using the Danger Assessment for Immigrant women (DA-I) [41], and supporting safety planning using a purpose designed booklet to guide discussion. The booklet and guided discussion were adapted from the DOVE intervention [42], one of few RCTs to find significant decreases in IPV, sustained over two years. Further details about the intervention are reported elsewhere (Authors, pending). This paper reports on a survey carried out three months after screening and intervention was instituted in the settlement services with the aim of describing the nature of abuse experienced, post-migration stressors and psychological distress among the sample, as well as actions taken by women after disclosing their experiences of abuse.

Methods

Study design

Our mixed methods study comprised a survey with women who attended study sites during the intervention period and focus groups / interviews with caseworkers. The survey was conducted approximately three months following participants’ visit to the settlement service. Our initial plan for a comparative study using interrupted time servies became unfeasible due to delays caused by the Australian international border closure during the COVID-19 pandemic.

Setting

Study sites were four government funded SETS services, based in metropolitan (3) and regional (1) centres in the state of New South Wales, Australia.

Participants

Included participants were (i) female; (ii) aged 18 years and over; and (iii) accessing a SETS program.

Recruitment

Women who accessed each of the study sites during the intervention period were approached privately in the waiting area by a member of the study team. They wereinvited to register initial interest in participating in a telephone survey on women’s health and safety and their experience of the service. At two sites, bi-lingual team members conducted the recruitment. Information and consent forms were provided in English, Arabic, Farsi, Chinese and Vietnamese. Recruitment occurred prior to meeting with a case worker who conducted the screening. Some of those recruited to the study did not meet with a case worker, for example some were attending solely for the purposes of obtaining forms from the front desk. Further details on the survey methodology are provided in (Authors, pending).

Survey items

Survey items included: non-identifying demographic data; post-migration stressors, psychological distress, experience of using the service; recollection of screening; services received and steps taken toward enhancing safety. The SAHAR IPV screening intervention focussed specifically on intimate partner violence. Recognising that other forms of family violence occur, the survey included a question on violence by other family members for those who identified IPV. Recognising that under-disclosure in response to screening occurs, survey participants whose responses indicated that they had not disclosed abuse in response to screening at the service, were asked the ACTS tool questions again, under survey conditions of anonymity. The ACTS screening tool asks women how often in the past 12 months a partner/ former partner has made them Afraid, Controlled them, Threatened or Slappeed/physically hurt them. A five point scale from never to very frequently is scored to yield a maximum of 16 with a cut off point of 1 [43]. Participants with no recall of being asked the screening questions were also asked the ACTS questions, recognising that some of the sample did not meet with a case worker on the day of presenting to the service.

Measures

The nature of IPV experienced was explored using the Composite Abuse Scale- Short form (CAS-SF) [44, 45], designed to measure prevalence of IPV. The short form CAS-SF includes 14 items and was selected over the full version in the interests of maximising total survey completion within a 25 min period. For the CAS-SF, if a case did not contain responses to at least 70% of sub-scale items it was counted as missing. Mean substitution was used for missing scores in cases where at least 70% of sub-scale items had responses; however, for the sexual abuse sub-scale, when responses were missing for either or both items, this variable was counted as missing (scoring guidance from the developers of the CAS_R-SF).

Distress was measured using the Kessler Psychological Distress Scale (K6) [46]. The K6 was scored using the Australian standard [47], with a five-level response where 1 was the minimum score for an answer (none of the time) and 5 was the maximum score (all of the time) with a minimum possible score of 6 and maximum possible score of 30. Those who did not answer all the questions were excluded from the analysis. Based on standard validation studies [48] and converted to Australian scoring, those with scores of 6–18 were categorised as having no probable serious mental illness and those with scores of 19–30 as having probable serious mental illness [47].

Post-migration stressors used the item from Building a New Life in Australia, a longitudinal study following humanitarian migrants from resettlement [27, 49]. The item gauges experiences after arrival including economic stress, English language barriers, family conflict in Australia, loneliness, discrimination, worries about family in home country and problems with adjustment to life in Australia. Survey participants who indicated disclosure of abuse at the settlement service in response to the screening questions, were asked about their experience of the response. Areas canvassed included: helpfulness of the caseworker, referral to the site’s IPV worker and provision of risk assessment and safety planning.

Sample size and survey administration

The study was designed to achieve a sample size of 396 to estimate prevalence of IPV.

Telephone surveys were administered by a team of bilingual research assistants with experience supporting refugee women in Arabic, Farsi, Dari, Chaldean, Assyrian, Mandarin, or Vietnamese. The national telephone interpreter service was used for participants requiring support in other languages. Full consent was taken at the time of the survey, after informing participants that the key focus of the survey was the IPV screening conducted at study sites at the time of their initial consent to re-contact.

Ethical considerations

The research was conducted in line with WHO ethical and safety guidance on conducting research with women who have experienced violence [50] For example, surveys were conducted only after ascertaining that participants were safe to proceed. A panel of former refugee women with direct/indirect lived experience of IPV guided the study which was approved by the University of (removed for blinding) Human Research Ethics Committee (2021/388).

Analysis

Identification of IPV was based on recall of responding positively during the site visit to the ACTS questions or administration of the tool in the survey, for those with no recall of being asked the questions. Statistical analysis was conducted using SPSS v.29. Descriptive data is presented for the Composite Abuse Scale results, pre-migration stressors and psychological distress (as measured using the K-6) and actions taken in response to disclosure. Cross tabulations were carried out for disclosure/non-disclosure of IPV against psychological distress; post-migration stressors; time since arrival in Australia; first/subsequent visit to the service; number of prior service visits, country of birth; language spoken at home and age. Chi square or Fisher’s exact test and odds ratio (95% CI) were used to determine associations between variables. Some participants did not answer all items, so totals vary between measures.

Results

During the four-month study period, 429 women visiting study sites were invited to be contacted for a survey on women’s health and safety, of whom 54 declined and 375 (87%) consented. At re-contact, 321 were successfully followed up and consented to participate in the survey (86% response rate). Another 32 women were unable to be contacted, 19 declined and three surveys were incomplete. Participants had a mean age of 44.3 years, came from 24 countries, most commonly Iraq (47%); Syria (13%); China (11%) and Afghanistan (10%) and spoke 25 different first languages. 69% had lived in Australia for 3–5 years. The most common forms of support being sought by study participants attending the SETS service on the day of recruitment were information/ advice/ referrals; group activities; advice on education/ training or employment and English classes. Over a third of participants (118; 37%) were visiting the service for the first time on the day of screening and recruitment, with 20% (65) having visited more than 10 times. Further details of participant demographics have been reported elsewhere (Authors, pending).

Identification of IPV

In response to the survey, 27 participants indicated they had disclosed IPV at the settlement service in response to the ACTS questions. A further thirteen participants reported they had not previously been asked the screening questions, but reported abuse against the ACTS tool for that time frame when asked these questions in the survey. An additional eight participants recalled being asked the ACTS questions, indicated they did not disclose abuse at the time of their visit to the site, although they reported abuse on the survey. Among those who recalled being asked the screening questions, the under-disclosure (false negative) rate was 4% (8/180). In total 48 participants identified current IPV (defined as past 12 months), or 15% of the whole sample who responded to the relevant survey items (48/316), with five women not responding to the ACTS questions in the survey.

Nature of abuse experienced

All women who identified current IPV were asked the CAS-SF questions about the nature of the abuse they had or continued to experience, with 45 participants answering enough items to score responses. By categories of abuse according to CAS scoring, 53% (n = 24) reported experiencing psychological abuse, 47% (n = 21) reported physical abuse, and 36% (n = 16) reported sexual abuse. As reported in Table 1, the most common types of abuse were: forced isolation from family/ friends (56%; n = 25); blame for abusive behaviour (53%; n = 24) and put downs (44%; n = 20). 29% (n = 13) experienced three categories of abuse and 18%, two categories of abuse (n = 8). Acts of abuse had rarely occurred only once, but more commonly daily or weekly (Table 1).

Table 1 Abuse types and frequencies: composite abuse scale-short form (n = 45)

In addition to abuse by their partner/husband, 34% (n = 16) of the women experiencing current IPV, additionally reported being “frightened, controlled, or hit” by other family members. No further data was collected on these experiences or the relationship to that family member.

Psychological distress

Among participants who answered the psychological distress items, 24% (n = 74) gave responses indicating probable serious psychological distress on the K6, with 11 10 responses missing. Women who indicated IPV were significantly more likely to experience psychological distress (OR 3.35 [95% CI, 1.71, 6.55]) with 46.5% scoring probable serious psychological distress compared to 20.6% of the non-abused sample (Table 2).

Table 2 Psychological distress and IPV disclosure (n = 305)

Post-migration stressors

Post-migration stressors were reported by all 316 participants. Analysis of the data including IPV status, indicates that multiple stressors were common, with 30% of the sample experiencing five or more different stressors (96/316). The most frequently reported stressor was English language barriers (85%, 272/316), followed by worries about family in their home country, economic stress and loneliness (Table 3). The post-migration stressors of discrimination, loneliness and family conflict were significantly higher for women who disclosed IPV, compared to those where abuse was not disclosed (Table 3).

Table 3 Experiences of post-migration difficulties (n = 316)

Further associations with IPV disclosure

Other associations considered in relation to differences between the group of women who disclosed abuse and those who did not. This included consideration of time since arrival in Australia. Based on chi-square test of independence, with all expected cell frequencies greater than five, no statistically significant association was found between disclosure of IPV and: time since arrival in Australia; first vs. subsequent visit to service; country of birth; language spoken at home or age.

Responses to disclosure

In answer to questions, about their experience of the response to disclosing abuse at the site 25 of 27 participants provided further information. It was not clear from survey responses which participants were referred to the onsite dedicated IPV worker.92% (23/25) of women reported the caseworker as very helpful with two reporting them very unhelpful. Of these two, one made a comment that the caseworker had suggested she seek a referral to a psychiatrist by her family doctor. This participant had previously accessed services in relation to IPV and was comfortable with being asked the questions. It appeared the second participant did not regard herself as having experienced abuse, stating “It was very helpful. I know who to call in case something happens.” 80% (20/25) of women responding to the question Did the worker arrange for you to get help from someone else? recalled the case worker arranged further assistance. Five women reported not receiving further help, with one woman saying she was helped by her contact with the case worker; three women reported not being referred to anyone, and one woman moved to another location further away from the original service. Results from focus groups held with case workers at sites which will be reported separately, indicate that many women declined offers of further referral (Authors, pending). Among the 20 women for whom further referrals were made, most commonly this was to a social work or counselling service (15), followed by legal advice (7), police, IPV specialist/ GP (6 each), housing/ emergency accommodation (4), child protection service (1).

Risk assessment using the DA-I was part of the SAHAR follow-up intervention, Participants were asked When you told the worker about being frightened, controlled or hit did she ask you questions about how much risk you might be at, or how much danger you might be in? with prompts of examples it would have been questions like - is the risk increasing? does he have a weapon? does he threaten the children? In response, 20/25 recalled being asked these questions, three reported they were not asked about risk and two did not recall. In relation to helpfulness of these questions, 18/20 (90%) reported the questions as very helpful, one woman each reported the process as neutral or unhelpful. Safety planning was reported by 19/25 women (76%), five did not recall and one said this did not happen. Seventeen participants (90%) reported the safety planning process was very helpful, with two participants of the 19 who responded to this item indicating it was “somewhat helpful.”

Eighteen of the 27 women (67%) who disclosed abuse during the screening process at sites reported taking one or more actions to improve their safety. These actions included acting on the referrals made by caseworkers: talking with a counsellor (13/18), reporting to police (9/18), getting legal advice (8/18), talking to a family member (7/18), taking out an apprehended violence order (6/18), talking with friends (6;33%), staying with another person for a while (4/18) and talking to a religious/ community leader or primary health doctor (3/18).

Participants who disclosed abuse at the settlement services were provided an opportunity during the survey to make comments about anything “good or bad” that occurred as a result of being asked the screening questions. Nine women provided brief comments, all of which indicated benefits from their experience of disclosing and receiving support:

I know a lot more information, and knowing where I can go to seek for help when needed.

I get so much help after speaking out.

I get the help I needed from the community.

I feel better.

Many supports whenever I need supports.

I have awareness.

I have more information.

I am feeling better.

While the proportion of women who accepted a referral to the dedicated IPV worker was low, as reported separately (Authors, Pending), it appeared that the assistance received was relevant and built safety.

Discussion

This study tested a culturally tailored intervention to identify and respond to IPV with women receiving settlement services. A three month follow up including the Composite Abuse Scale-SF found that the the most common form of abuse was psychological, followed by physical and sexual. Similar to other applications of the CAS-SF, the most commonly reported abusive acts were isolating the woman from family/friends, blaming her for his abusive behaviour and put downs [51].

Relatively high levels of psychological distress and post-migration stressors were reported by all women, but these were significantly higher for women who had experienced IPV. The majority of women who disclosed abuse to the service found the caseworker’s response to be very helpful and usually involved risk assessment, safety planning and referral for further assistance. It is widely accepted that many women asked IPV screening questions in health or other settings elect not to disclose abuse even if it is currently being experienced, due to fear, shame or mistrust [52, 53]. Recognsing this, most protocols for screening embed the offer of a discreet wallet sized information card, regardless of whether abuse is revealed. The measurement of the rate of “false negatives” among cohorts of women asked screening questions is a valuable reminder that not all women feel safe enough to disclose and that screening is not a prevalence measure. In this study, 4.4% of women indicated in their survey responses that they had elected not to disclose the abuse they were experiencing when asked screening questions. This is much lower than found in other research, with one study finding 14% (34/240) of women screened in health services gave “false negative” responses [54] and emergency department research indicating up to 80% of abused women screened in that setting elected not to disclose [55].

Among the women who indicated current IPV, one third reported concurrent abuse by other family members. This question was only posed to women who had already indicated experiencing IPV. Abuse by other family members appears higher in this cohort than in some other samples. For example, a large representative Australian prevalence study which found that 23% of women had experienced IPV, identified that 8% had experienced abuse by other family members [56].

Our findings on psychological distress and post-migration stressors confirm that both of these are high among refugee women in general. It is likely that the psychological distress and post-migration stressors which were significantly higher for women who reported IPV, were a result of the abuse experienced, though causation cannot be assumed. The significant association between IPV and family conflict and loneliness are unsurprising, however discrimination is more unexpected and suggests that those experiences of marginalization in society may contribute to occurrence of IPV. This result is a reminder of the complex inter-relationships between IPV and other stressors and the fraught nature of complexity of many women’s lives.

As reported, we found no association between length of residency in Australia and disclosure of abuse, with women who had arrived less than two years earlier being as likely as those who had been in Australia five years or more, to report abuse. On the one hand, those who recently arrived may be more likely to have experienced recent abuse due to exacerbating stressors of forced migration and/or the conflict which typically precedes it [57, 58]. There is also some evidence that abuse can increase post- arrival due to acculturation stressors, language barriers, non-citizen status, economic insecurity and limited social support [7]. On the other hand, it is likely that time post-settlement brings more stability and safety to disclose abuse, leading to higher rates of disclosure. Further research could establish whether there is an actual change in the levels of abuse experienced in immediate post-arrival period compared to longer residency.

Similarly, we found no association between likelihood of disclosure at first or subsequent visits to the settlement service. This is counter-intuitive, as it might be assumed that the multiple visits to a service build trust, which increases the likelihood of disclosing IPV. However other research similarly found that many women disclose abuse on their first visit to a service. For example a statewide health program in antenatal, mental health and substance abuse services, with a protocol for IPV screening at first visit, found that 23% (27⁄120) of those who disclosed abuse, reported that their response to screening was the first time they had told anyone about being hurt by their partner [54]. Qualitative research also supports that women make rapid decisions to disclose experiences of abuse to health workers who are seen as caring [59,60,61] and that immigrant women are deterred from disclosing IPV by unsupportive staff within mainstream services [62]. These strands of evidence suggest that settlement services are viewed as safe places to disclose and receive support by many women, reinforcing the potentially important role of screening at these sites.

The most recent United States Preventive Services Task Force [32] review of evidence for IPV screening with women of reproductive age, found a net health benefit, particularly where ongoing supportive intervention is provided. Other evidence suggests that screening should be limited to groups at higher risk [63, 64]. Our intervention was modelled on the DOVE study, recognising that women who disclose abuse through screening require follow-up, utilizing a considered and planned approach. The results suggest that settlement services are well placed to provide both screening and response to IPV disclosures, having the requisite skill sets to identify, respond and refer appropriately.

Limitations

We had initially planned an interrupted time series design but this was altered due to the cessation of refugee intakes during the pandemic. Not having a comparison group for testing IPV screening and intervention is a study limitation.In addition, participating services may not be typical of all settlement services; the motivation to improve responsiveness to IPV was high among those participating in this study but might be less so in other settings. We recognise that collecting survey data approximately three months after women’s experience of screening and response may introduce recall bias.

Implications for research, policy and practice

Further research is warranted with women who disclose abuse, providing follow up over a longer timeframe to understand impacts from the intervention. Acceptability to service users, an essential criterion for introducing any screening test [65], is reported separately (Authors, in press). High rates of post migration stressors among refugee women who identify IPV point to the need for wholistic interventions which address their marginalized status and address the barriers this research points to in terms of their access to mainstream services addressing IPV. Exploration of the role of discrimination in particular is warranted given our finding of its significant association with IPV.

This intervention identified serious forms of IPV among resettled refugee women, and was well received by these women in this study highlighting the strengths of culturally tailored responses. As a group who face many barriers to accessing mainstream services [18, 20], the SAHAR intervention shows promise as a means to addresss abuse. Social workers in all jurisdictions are called on to respond to intimate partner violence and vulnerable population who experience it. Introduction of policies which encourage or require settlement services to introduce mechanisms to identify intimate partner iolence and to fund dedicated on site IPV workers would be a valuable investment for governments.

Conclusion

The SAHAR study demonstrated that IPV screening and a response entailing risk assessment and safety planning that is culturally tailored and supported through training and translated tools, can work effectively to support former refugee women. Findings of higher rates of psychological distress and post-migration stressors indicate the importance of support for this group, most of whom in this study, through the screening and response intervention were able to access support from caseworkers, risk assessment, safety planning and appropriate referrals. Settlement services are an important resource for those who arrive through forced migration and intervening to address IPV has potential to strengthen the value of such services.

Declerations