“Made Me Feel Connected”: A Qualitative Comparative Analysis of Intimate Partner Violence Routine Screening Pathways to Impact

Abstract

Intimate partner violence (IPV) routine screening is widely implemented, yet the evidence for pathways to impact remains unclear. Of the 32 abused women interviewed 16 weeks after antenatal IPV screening, 24 reported positive impact, six reported nil positive impact, and two reported negative impact. Using qualitative comparative analysis (QCA), key conditions for positive impact were care in asking, and support and validation from the midwife. Lack of these and lack of continuity of care were relevant to nil positive impact. Benefits included naming the abuse, connection, unburdening, taking steps to safety, and enabling informed care. Disclosure was not required for positive impact.

Keywords

intimate partner violence domestic violence routine screening routine inquiry QCA

Introduction and Background

Between 38 and 59% of women presenting to health care providers have experienced lifetime intimate partner violence (IPV) based on a systematic review of 37 studies (Sprague et al., 2014), with women who experience abuse using health care services more frequently than women who do not experience abuse (Kothari et al., 2015; Zakrison et al., 2017). Despite this, health care systems trail behind other agencies in responsiveness to IPV (Garcia-Moreno et al., 2015). As a result, opportunities may be lost for early intervention. Because social isolation is common among women who experience IPV (Stark, 2007), health service visits can provide opportunities for support, reinforcing the importance of strong health system responses. In addition, women subjected to IPV report that health care providers are the professionals that they trust with disclosure of abuse (Feder, Hutson, Ramsay, & Taket, 2006).

We use the World Health Organization’s (WHO; 2013, p. vii) definition of IPV as “behaviour by a current or former intimate partner causing physical, sexual or psychological harm which may include physical aggression, sexual coercion, psychological abuse and/or controlling behaviours.” Routine screening for IPV has been widely introduced in health systems in many countries in an effort to increase its identification (O’Doherty et al., 2015). Although widely referred to as “screening,” the term routine enquiry may be a more accurate descriptor, given that there are potential “intervention” effects of this practice (Klevens & Saltzman, 2009). Despite its widespread implementation, there has been ongoing argument that there is insufficient evidence documenting positive outcomes of universal screening (O’Doherty et al., 2015). Results from large randomized control trials (RCTs) have found no significant differences in routine screening intervention on most outcome measures (e.g., Hegarty et al., 2010; Klevens et al., 2012; Koziol-McLain et al., 2010; MacMillan et al., 2006; MacMillan et al., 2009). The struggle to find interventions which reduce IPV is not limited to screening, but there is a lack of clear evidence for other types of interventions as well, such as intensive advocacy (Rivas et al., 2015). RCTs have faced challenges in finding differences between control and intervention groups due to difficulties in identifying a sample without asking about abuse, high study attrition rates, and potential Hawthorne effect from repeat measures (Spangaro, Zwi, & Poulos, 2009; Wilbur, Noel, & Couri, 2013). However, other research indicates that abused women value the opportunities provided by routine screening to discuss their experiences and access support (Bacchus et al., 2016; Koziol-McLain, Giddings, Rameka, & Fyfe, 2008). There is a need to bridge these disparate findings. Given evidence of underdisclosure of IPV (Evans & Feder, 2016), many women’s views are not heard in this research. It is perhaps assumed that women who do not disclose their abuse will not experience any impacts from the screening process. Research is needed which explores the impact of IPV screening intervention for women who elect to disclose their experiences of abuse to health service providers, as well for those who do not. The complexity and diversity of women’s situations point to the need for nuanced understandings of how positive and negative impact occurs, the conditions and circumstances that influence these outcomes, for whom these occur, and, most importantly, to identify the factors that are necessary and sufficient to lead to positive impact. Recognizing that different pathways may lead to the same outcome (i.e., positive impact), research that informs ways in which interventions can be targeted to the diverse needs of women is needed.

Earlier research by team members about women’s experiences of routine screening found that some women reported valued impacts from screening (Spangaro, Zwi, & Poulos, 2011). This resulted in a tentative model for women’s perceptions about the impacts of screening, which included naming the abuse and gaining a sense of connection (see Figure 1). The study showed that benefits were not restricted to those who disclosed their experiences.

Figure 1.

Model for women’s perceptions about the impact of screening.

The current study aimed to refine and extend this model and understand the pathways leading to perceptions of positive impact of screening and, equally, pathways leading to perceptions of neutral or negative impact (hereafter termed situations of “nil positive impact”). It follows our earlier papers reporting pathways to women’s decisions to disclose or not disclose abuse in response to routine screening, which drew on the same sample (Spangaro, Herring, et al., 2016; Spangaro, Koziol-McLain, et al., 2016). We included women who disclosed their experience of IPV to a midwife and those who elected not to. We asked women about their perceptions of impact from routine screening and the health service response. Acknowledging that, in Australia, Aboriginal women are disproportionately affected by IPV (Willis, 2011) and experience health systems differently, we oversampled Aboriginal women. These findings have been published separately (Spangaro, Herring, et al., 2016; Spangaro et al., 2019) to allow sufficient reporting of our participatory approach; the divergent findings; and specific implications for research, policy, and practice. Although the impacts we identified among Aboriginal women were similar to those of non-Aboriginal women reported here, and many of the conditions were common, a key difference was the centrality of the concept of cultural safety (Spangaro et al., 2019).

Using a qualitative configurational approach, we determined the pathways to women’s perceptions of positive or nil positive impact, taking into account the different experiences of disclosing and nondisclosing women. Using thematic analysis, we also unpacked the elements of positive impact to identify discrete benefits women articulated. We also considered explanations for absence of positive impact.

Method

Methodological Approach

We employed qualitative comparative analysis (QCA), a structured method developed for the social sciences to study complex phenomena (Ragin, 2014), to test the IPV screening impact model and map conditions, leading to the screening outcomes. QCA can “drastically shorten the distance between qualitative and quantitative methods, [. . .] By translating qualitative data, including potential causal factors, into numerical format and systematically analysing it, [. . .] allowing causal patterns in the data to be found” (Befani, 2016, p. 5). The method recognizes that outcomes commonly result from multiple “conditions” and that different combinations of such conditions can produce the same outcome (Berg-Schlosser, de Meur, Rihoux, & Ragin, 2009). Analysis results in identification of one or more “pathways” or combinations of conditions, leading to the outcome of interest. QCA preserves case diversity and complexity, distinguished from regression analysis through considering the collection of conditions as a whole and in creation of multiple pathways (Befani, 2016). Analysis involves reduction of potentially relevant conditions to binary measures, where 0 = low/weak/absent and 1 = present/high. After case-by-case coding using an agreed data dictionary, the impact of a condition’s presence or absence in combination with other factors is systematically examined by means of a two-dimensional matrix (Rihoux & Ragin, 2009). While seemingly reductionist, QCA is iterative, relying on holistic understanding of cases and the relationship between cases and theory. This holistic understanding is useful, given the complexity of abused women’s interactions with health services, in light of the shame, coercion, and risks they often experience. Previous QCA applications include exploration of child protection decision making (Hackett & Taylor, 2014) and factors important to neighborhood recovery following Hurricane Katrina (Jordan, Javernick-Will, & Amadei, 2014).

In this analysis, we map pathways to both the outcome (women’s perceptions of positive impact from screening) and absence of the outcome (nil positive impact), resulting in two solutions each with multiple pathways to the outcome. The IPV screening intervention was considered to include both the asking about recent experiences of IPV and the associated health care response in relation to any disclosure. Data were principally elicited through a question in the study interview guide: “For you, was being asked those questions in the health service a good thing? A bad thing? Not sure? Can you tell me more about that?” For responses to be coded for positive impact, participants needed to identify personal benefits, not simply agree with routine screening.

We also employed thematic framework analysis of the transcripts, which allows classification and organization of key themes, concepts, and emergent patterns (Ritchie & Spencer, 1994). This enabled us to consider the different benefits, expanding on the perceived impacts in the original model. Specification of the types of impact identified in the thematic analysis was termed “benefits,” to distinguish them from the outcomes of the QCA referred to as positive and nil positive “impact.”

Study Locations

Our study sites were three antenatal clinics in Sydney, Australia, where state health policy requires routine screening about IPV on entry to antenatal services (NSW Department of Health, 2006). According to the policy, two scripted questions – Within the last year have you been hit, slapped, or hurt in other ways by your partner or ex-partner? and Are you frightened of your partner or ex-partner? – are followed by the offer of an information card regardless of the answer. Women who disclose are offered referral to the social work team or other relevant specialist IPV agency as indicated. The state-wide policy is supported by staff training, annual monitoring, and localized procedures (NSW Ministry of Health, 2016). IPV screening is incorporated into the assessment at women’s initial visit, generally 10-16 weeks gestation. State-wide monitoring data indicate that 90% of antenatal patients are asked the questions, with 3.4% disclosing IPV (NSW Ministry of Health, 2016). Clinics were selected to reflect the state’s ethnic and socioeconomic diversity. Interviews were conducted approximately 16 weeks after the routine screening intervention.

Inclusion Criteria and Recruitment

Participants were English speaking, 28+ weeks pregnant, and recalled being asked the IPV questions at their booking-in visit (average 16 weeks earlier). All participants had experienced past 12 months or current IPV or fear of a partner at the time of screening, as assessed by research assistants (RAs), who approached patients at a regular antenatal visit and invited them to a private room. For safety reasons, the study purpose was initially masked, with additional psychosocial questions included in the eligibility screen. Women who disclosed IPV to the RA using the same questions as used in the initial screen were then informed about the focus of the study and invited to participate in a single, same-day interview. Recruitment occurred during weekly half-day clinic visits over a 14-month period.

Data Collection

Semi-structured face-to-face interviews taking an average of 45 min were audio-recorded with participant consent. Participants were provided AUD$30 in acknowledgment of their time and contribution. The interview schedule guided exploration of experiences of the IPV questions, decisions to disclose, perceptions about the impact of screening, receipt of information or referrals, prior disclosures, statutory agency contact, and current status of relationship with the abusive partner. Interviews were transcribed, and cases were assigned a unique identifier and pseudonym. Women gave consent for one-off access to their medical notes for data on actions taken.

Ethics

A reference group comprising senior representatives from participating sites and the health department’s IPV policy officer advised the study team. Study procedures were consistent with ethical and safety guidelines for research on IPV (Watts, Heise, Ellsberg, & Garcia Moreno, 2001) and approved by the South Eastern Sydney Local Health District (SESLHD) Human Research Ethics Committee (Ref. 12/191).

Data Analysis

Transcripts were independently binary coded by two pairs of team members (for condition presence or absence), with team discussion to resolve discrepancies. Iterative analysis involved close reading of cases to identify perceptions about the impact of screening, whether positive, negative, or neutral. Data were then analyzed using fsQCA software (Ragin & Davey, 2014) to provide pathways to positive impact or its lack. Thematic data were initially coded using NVivo 10 by J.S., with the team as a whole agreeing on definitions and elements of the model.

Specification of Conditions

QCA starts with specification of conditions. Table 1 lists the potential conditions with examples of coding for condition present or absent, showing shading on conditions that remained in the final model solutions. The conditions of Care and Adverse statutory contact were identified inductively during analysis of data, with all other conditions explicitly asked about in the interview. The original model and findings (Spangaro et al., 2011), derived inductively from interview themes, formed the basis for the conditions to be tested in the analysis (Conditions 1-6 and 8 in Table 1). Four further potential conditions were identified during study design (9-12) and two additional conditions identified by the team during the final analysis (7 and 13).

Table 1.

Conditions Tested for a Positive Impact of IPV Screening Intervention (N = 32).

Condition name	Definition	Cases present	Exemplar text when condition present (code = 1)	Exemplar text when condition absent (code = 0)
1. Disclosure	The woman reports that in response to screening questions she disclosed past 12 months of abuse or fear of a partner/ex-partner	24	“I just told her the truth. That I had been in a DV relationship, still am.”	“So I just say to her no, nothing like that.”
2. Care	The woman perceives the midwife is caring, that is, she shows interest and a nonjudgemental attitude in the way she ASKS about abuse.	20	“She explained all the questions to me, and she took the time and actually listened to what I was saying and had like, a conversation with me.”	“Her body language itself didn’t feel like she wanted to be there, and rushing through the questions; yep, yep, yep, rush through, rush through.”
3. Abuse not current at time of research interview	The woman’s response indicates she is no longer experiencing hitting/hurting or fear of partner at the time of the research interview (including not being exposed to postseparation harassment/coercion).	11	“I don’t have anything to do with him.”	“He’s cut back on hitting me a lot.”
4. Prior disclosure	The woman has named the abuse to another person (prior to screening), including family or a service.	24	“Yeah, my family knew. Everybody.”	“’Cause I hadn’t really told anyone before, so, yeah.”
5. Prior service use	The woman reports contact prior to the screening with support services to address IPV.	11	“I was seeing a counsellor already.”	“No, like not even my family were aware of it.”
6. Prior statutory contact	The woman reports prior contact with police or child protection statutory agency.	17	“My mum rang the police for me.”	“No, it was never anything bad enough.” (For police/child protection involvement)
7. Adverse prior statutory contact	Subgroup of Condition 6 where a woman describes contact as indifferent, humiliating, or otherwise negative.	3	“I called the police, but then they always say, ‘It’s your word against his. We’ve got no witnesses’.”	“I was really anxious about a report to [child protection] but it turned out to be a good thing.”
8. Support and validation	The midwife offered support and validation in her response to disclosure (focus is on midwife’s actions/behavior).	18	“You could feel she was interested in the story and how you feel and how to make sure you have all the support you need.”	“I said yes, she said, okay and then she just turned around to the computer and asked the next question.”
9. Received card	The woman received information card in association with screening.	19	“Yeah, purple card, exactly, just in case.”	“No, I haven’t received this.”
10. Used card	Receiving the card was useful, for example, it resulted in identifying new information about IPV or led to new understanding/naming of abuse.	7	“The information in it, that was good, about how men control your money and by isolating you.”	“I haven’t looked at it.”
11 Services offered	In association with screening, a woman was given information about or offer of referral to services to address IPV.	23	“Just giving me options, like if I want to see a counsellor here or if there was any other services she could refer me to.”	If no disclosure, not asked and coded 0
12. Services used	The woman accessed services as a result of disclosure at screening. Includes clinic social worker or IPV service.	12	“Yes I saw the social worker. She was really helpful and I saw her throughout, plus after when I had the baby.”	“She said she can refer me to the social worker but I said I was okay at the moment.”
13. Continuity of care	The woman continued to receive her antenatal care from the same midwife who asked the screening questions through the pregnancy at all or most visits.	7	“And she was with me all the other visits afterwards.”	“I opened up to her a lot but I haven’t seen her again after that visit.”

Note. IPV = intimate partner violence.

QCA

Analysis using QCA involves identification of new conditions, replacing some conditions with others, and reexamining in a qualitative and “thick” way the cases involved (Rihoux & de Meur, 2009). Data tables were constructed with a row for each case containing the binary coding for the conditions of interest (Rihoux & Ragin, 2009). Primary metrics provided by the fsQCA software analysis are consistency, the degree to which cases with a given set of conditions exhibit the outcome—in this case perceived positive impact—and coverage, the degree to which a given pathway explains the cases analyzed (Jordan et al., 2014). Typically, abundant condition options are present, but the optimum number for causal modeling in a sample of 10-40 cases is 4-7 (Berg-Schlosser & de Meur, 2009). Patterns of conditions were systematically tested to identify model solutions with high coverage and consistency, using a consistency cutoff of 0.8 for inclusion of configurations as recommended by Ragin (2008). Solutions contain no contradictions; that is, the same configurations of conditions cannot lead to both the outcome and the absence of the outcome. In addition, in identifying best fit, we aimed for solutions which made theoretical sense, aligned with narratives identified in the cases, and were parsimonious, that is, solutions that are as simple as possible but still allow for causal complexity (Berg-Schlosser et al., 2009). Cases can fit multiple pathways; however, the parsimony principle also calls for solutions maximizing unique cover.

Findings

Sample Characteristics

Thirty-two women’s stories were included in the analysis (for filtering details, see Spangaro, Koziol-McLain, et al., 2016). Twenty-four of the women disclosed their experiences of IPV to the midwife in response to screening questions, and eight chose not to disclose their experiences of abuse. Sample characteristics are reported in Table 2.

Table 2.

Characteristics of Study Sample.

Characteristic	Abuse disclosed (n = 24)	Abuse not disclosed (n = 8)	Total (n = 32)
Average age at interview, years (range)	28 (17-41)	27 (21-36)	28 (17-41)
Average weeks pregnant at screening (range)	18 (11-32)	18 (13-22)	18 (11-32)
Average weeks between screening and interview (range)	16 (2-23)	15 (10-19)	16 (2-23)
Abuse current at the time of screening	19	6	25
Abuse current at the time of interview	15	4	19
Language other than English spoken at home (Arabic, Bengali, French, Hindi, Indonesian, Italian, Japanese, Macedonian, Samoan, Tagalog, Thai)	9	2	11

QCA—Retention of Conditions in the Model

Case and cross-case analysis led to a number of analytic decisions. Service Offered was excluded because of the 24 women who disclosed, all but one (Bridget) were offered referral to a service. Analyses containing Prior Service Use, Prior Statutory Contact and Adverse Prior Statutory Contact, and Used Card did not differentiate those who perceived benefit from screening to those who did not, producing solutions with numerous contradictions. Although Prior Disclosure was a helpful construct for differentiating six women who had not previously told anyone about the abuse, which is a target group for service providers, it did not predict impact. The best model solutions without contradictions, fewest pathways, and most unique coverage contained the conditions: Disclosure, Care, Support and Validation, Abuse Current at Interview, Got Card, Continuity of Care, and Service Used. Final data tables and outputs are available from the authors.

Pathways to Positive Impact From Screening Intervention

Recognizing that reactions are complex, the women in this study were nonetheless clear in identifying whether, for them, screening was “on the whole” positive or negative. Positive impact, as defined in the “Method” section, was reported by 24 of the 32 women. Of these 24 women, 23 were offered a specialist service, but only 12 actually took up this option. The specialist service included seeing a social worker attached to the clinic and in some cases receiving further referral to external services. Accounts of women who declined the offer indicated that the main reason was a belief that a service was not necessary at that time, possibly related to the absence of ongoing abuse. Two women reported accepting follow-up by the social work team, but not ever being contacted. Surprisingly, Service Used and Continuity of Care both dropped out of the models.

The four pathways to positive impact are mapped in Figure 2. The solution coverage and consistency were both 1.0, indicating nil contradictions; 20 cases were covered uniquely, that is, fit only in one pathway (bold text in Figure 2).

Figure 2.

Pathways to positive impact from routine intimate partner violence screening intervention.

Pathway I in Figure 2 indicates that disclosure was not required for women to perceive positive impact from being asked about IPV. We titled this pathway Silent but Receptive. Both women (Clare and Amber) continued to experience abuse at the time of the research interview. Clare explained that she did not disclose because she was already receiving help and “a lot of it was in the past.” Pregnant with twins, Clare received the information card, which she reported was useful. Amber’s abuse included sexual violence, and she reported that although it could be upsetting when the abuse was raised, it was helpful for her to be asked.

Pathway II covered the largest number of women (15), all of them uniquely, so was the dominant pathway for positive impact from screening. It testifies to the prominence of Care (asking about the abuse with interest and no judgment) and Support and Validation (the midwife’s response to disclosure) in bringing about positive impact. These related but separate processes were both central for most women to perceive positive impact from being asked the screening questions, and apparently more important than receiving a specialist service or continuity of care. Pathway II included women who continued to experience abuse at the time of the research interview and those who were safe. We titled this pathway Supportive Asking and Response.

All of the women in Pathway III had prior contact with statutory agencies in relation to their abuse experiences, either police or child protection; intervention for abuse was not a new experience for these women. Despite formal agency intervention, four of six women continued to experience abuse. We titled this pathway Supported with Information to reflect receipt of the information card as its distinguishing condition:

I think it’s just the fact that when she gave me that purple card at the end, and then I realized that there’s so many other people out there who can help us in relation to our past experience. (Bridget)

It is clear from Bridget’s words that the provision of written information on IPV, and perhaps the universality of asking about abuse, signals that IPV is widespread and health services are ready to respond, perhaps contributing to a shifting of social norms in relation to IPV. A high degree of crossover was found between Pathways III and IV, with four women covered by both. As with Pathway III, the women in Pathway IV all had prior statutory agency contact. However, this pathway was distinguished by the condition of Abuse Current at Interview. For these women, screening may provide further opportunities to gain information and support. Pathways III and IV also contained the absence of Care as a condition. Rather than interpreting this as a requirement that the midwife be uncaring, we conclude that for these women positive impact was perceived, despite the lack of care when the questions were initially asked by the midwife. As Maya stated,

Because it’s been a while since I opened up to somebody about that, so it was good just when she just asked me. Actually, [I] just told it all, even if she wasn’t interested. I don’t know why. But it felt good to get it off my chest once I started.

Maya continued to live with her husband, relaying that he no longer punched her, but still hit the table and frightened her when under the influence of alcohol. This was despite the action she had taken to address the situation, involving both police reports and court appearances. Her situation is typical of many women who have taken steps to address the abuse but find that it continues. As Maya reported, “He is no bad man when not drinking.” We titled this pathway Still Looking for Solutions. These last two pathways highlight two issues. The first is that women’s situations are complex, with abuse often continuing despite intervention, and women often wishing to preserve their relationships. Second, women can be appreciative of health workers addressing IPV, even when responses are not well executed.

Pathways to Nil Positive Impact

Eight women identified nil positive impact from their experiences of screening. Of these, two women (Vani and Pippa) identified negative experiences from being asked the screening questions. These experiences are further explored in the thematic analysis. The remaining six women experienced neither positive nor negative impact but simply did not experience positive impact. Two pathways to nil positive impact emerged using QCA as shown in Figure 3. The condition Abuse Current at Interview dropped out of the model. Again, solution coverage and consistency were both 1.0.

Figure 3.

Pathways to NIL positive impact from routine screening about IPV.

Lack of Service Use and lack of Support and Validation are logical consequences when women do not disclose their abuse, however their absence, as well as lack of Continuity of Care was a factor for disclosing women as well who reported nil positive impact. Referring to the importance of validation, one woman recommended that health workers go so far as to “pretend” to be caring and compassionate, even if they don’t feel it.

The lack of Continuity of Care by midwives between antenatal visits was also an issue which women noted, as Bobby articulated:

Cause you get one lady that asks you about that type of stuff and then you talk to that person and you go to your next visit but THAT lady has no idea and then you go to your NEXT visit and then THAT person has no idea. Yeah. It kind of loses track of why the lady asked the first time.

Pathway I comprised women who had elected not to disclose their abuse. For this reason, we titled it IPV Not Considered, as abuse was not factored into the health service response. For those in Pathway I, it appeared that the lack of the information card again made a difference, as with the corollary benefit its provision made to nondisclosers with positive impact. Routine screening was not a negative or harmful experience for the women in Pathway I; rather, it was unremarkable. For a number of women, such as Nisha, the relationship with the abuser had ended, and although they may still be experiencing harassment or fear, it was seen as something in the past. In response to being asked whether the screening questions made any difference, she stated,

No probably not, because I’m one just to move on and deal with it and just worry about my kids. (Nisha)

Women reporting lack of impact were more likely to have been asked the screening questions in earlier pregnancies, made multiple prior disclosures, and used multiple services. Some of the women were also concerned about how the information would be used:

It didn’t achieve anything, because I didn’t tell them the truth. I think if they had asked the questions differently or given more information on why they wanted to know and what would be done with the information, people might be more inclined to tell. (Nell)

Pathway II which we titled Lacking Support and Validation included Pippa and Vani. Both women disclosed and reported that midwives showed Care in how they asked about the abuse and they were offered services. However, from their accounts, responses to their disclosures lacked Support and Validation. Pippa, who reported that she often felt sad during the pregnancy, explained,

She just gave the card. If I feel like—like if anything happens again to ring that number.

Both Vani and Pippa remained with their abusive partners and described relationships characterized by emotional abuse.

Ambivalence in Relation to Effects—The Good and the Bad

The nature of QCA is premised on a binary classification for outcomes as present or not present; in this case, women’s experiences are positive or not positive. The distinction was clear in the data; however, we acknowledge that alongside these outcomes, women’s perspectives were complex. Some women reporting positive impact also identified elements that were not helpful, exemplified by April:

At the time, I was probably scared and sort of really anxious about what’s going to happen now that I’ve said all these things and they’re going to make a report to (child protection). You know but it’s turned out to be a good thing that, yeah.

Following a report to the child protection agency, April and her partner were referred to an early intervention program for children at risk:

They’ve really taken the time with him to sort of break it down that it is abusive and its domestic violence, it’s not good for the kids. And he’s really taken it on board.

Conversely, some women with nil positive impact noted aspects of the process they appreciated. Although Rose did not identify any positive impact, she appreciated the midwife’s approach:

She was nice. She wasn’t rushing me. She wasn’t making me feel like I had to answer.

Unpacking the Positive Effects

To further understand the nature of the binary impacts identified in this second part of the article, we turn our attention to the thematic analysis and fit to the original model. From close reading of transcripts, thematic coding in NVivo, and team discussions, we refined the original model to include five domains of benefit for women having experienced positive impact. The domains included Naming the Abuse, Connection, Unburdening, Steps to Safety, and Enabling Informed Care.

Naming

Naming abuse is defined both in our original model and here as occurring when, because of routine screening and the service response, women reached a new understanding that their experiences, past or present, constitute abuse. This benefit was identified in our original model. The analysis in this study indicated that it proved a core benefit with this sample, reported by 17 of 24 women. For some, this involved identifying for the first time that their experiences constituted abuse. For others, realizations were subtler:

And then the midwife started talking to me. [. . .] And I didn’t realize that all that was happening to me was actually domestic violence. (Della)

It made me realise the fact that I had—gone through this.—it was a lot worse than what I thought it was at the time. (Farlie)

Clare and Amber (Pathway I for positive impact) both reported Naming. Neither woman disclosed their abuse. It seems that Naming, which is an internal process, may be a primary way routine screening can assist women who chose not to disclose their abuse:

And it just makes you think, is this happening to me? Do I have a fear? Is there domestic violence here? (Clare)

It does make you stop and think about what you’re going through and what it is and what is wrong and how much you can deal with. (Amber)

Most of the women in the largest pathway to positive impact (Pathway II; 12/15) experienced Naming. Presence of the condition No Prior Use of Service seemed particularly linked to Naming, with 13 of 21 women who had not previously used a service reporting Naming, suggesting this first encounter with a professional in relation to abuse unlocked this benefit.

Connection

Connection was also identified in the original model, defined as a sense of backup and decreased isolation as a result of the response to the woman’s disclosure. The construct retains two of the three elements of Connection from the original model, that is, Sense of Back Up and Felt Supported. Sixteen women identified Connection as a benefit of the experience of the screening intervention. All had disclosed their abuse. The value ascribed to having “backup” was clearly evident in the data:

I think it sort of empowers you, because you do feel like you have back up, it’s not just my opinion against his anymore. (Sophie)

One woman was explicit in identifying this as a sense of connection—which may counter the social isolation created by the abusive partner’s coercion:

But also her personality really helped, because she was really mild and very funny and we talked about it. So she actually made me feel connected. (Dee)

Again, most of the women in positive impact Pathway II (12/15) were coded for Connection as a benefit from screening, with no consistent pattern for the other three pathways. This suggests that this benefit derives strongly from the midwives’ care in asking about abuse and the offer of support and validation in response to disclosure.

Unburdening

We understood Unburdening as a sense of release or reduced distress attributed by women to sharing the experience of IPV with the midwife. Sixteen women were coded as receiving this benefit, all, logically enough, women who had disclosed:

Just being honest, it’s like a feeling that you’re not stressed about it or anything like that and just having told her, it was a good thing. (Karen)

I just told them. And I feel better if I share with somebody. I just feel better. (Nisha)

It is a good thing because some womans still not ready to open up to people about their past. So it—it will help to release the pain. Yeah, it will help. (Rogie)

Prior service use and whether the abuse was current were not linked to Unburdening among this group of women, suggesting that Unburdening was valued by women in a range of circumstances and that pain from the abuse continued beyond its cessation. Again, most of the women in Pathway II (12/15) were coded as Unburdening being a benefit.

Steps to safety

This benefit described by 19 participants combines Narratives of Competence from the original model with other actions women took that indicated their agency, such as talking to family members to strengthen the support system and challenging their abusive partner:

It opened it up for conversation and I started saying more things that had happened. I suppose my family at the time thought he was just calling me ugly, kind of thing. It was nothing LIKE that—so then they went “Ohhhh. Okay.” (Farlie)

I told [his friend] that if he ever comes around near my house, or my ex-partner comes to my house, the police will be called. (Sandy)

Such responses were not limited to those who had disclosed the abuse. Clare described how the experience of being asked about the abuse made her relive some frightening memories, yet also prompted her to recognize her resourcefulness:

You see how far you’ve come and see the milestones that you took to get there. (Clare)

Taking Steps to Safety was also not limited to those who had received a specialist service. The steps some of the women took may be considered small but nonetheless should not be discounted. As Emelia recounted,

The social worker asked me why I was still in contact with him, or if I was waiting for something from him and I said “Yes, just because I’m not ready to do anything more than this.” It’s like—your personal—your emotional strength, you got like a limit and you just take care of yourself and the pregnancy.

The complexity of women’s situations in relation to experiences of abuse needs to remain in sight, particularly during pregnancy, so that small acts of resistance are acknowledged.

Enabling informed care

Present for 23 of 24 women who reported positive impact, this benefit was classified as being present when the woman perceives that through disclosing the health service is enabled to provide more appropriate care. This included, but was not limited to, a better understanding by the health service of the woman and her child’s health needs, provision of options to the woman, and documenting the abuse in the medical record in case of future events:

It’s helpful because if it happens again, at least it’s on the record. (Ashling)

Then we got in contact with the other services [. . .] and things started happening. (April)

Unpacking Perceptions of Nil Positive Impact

Neither positive nor negative

As explored earlier in the pathway analysis (Figure 3, IPV Not Considered Pathway I), six of the 32 women identified neither positive impact nor actual harm from being asked about abuse. All six women had elected not to disclose their experiences of abuse.

Distress uninterrupted

Two women in the sample identified experiences which, while not quite adverse, were negative, reflecting lost opportunities to reach out to women who were both still in abusive relationships at the time of asking. Pippa and Vani who also shared Pathway III for nil positive effect elected to disclose, but neither experienced health service responses that acknowledged their distress:

I told her and after that I didn’t see her again. And the doctor didn’t ask anything about this. [. . .] I just trying to forget. The thing is it was quite hard to forget, but still I tried to forget, because if you can’t forget everything it will be hard to go on. (Vani)

Pippa indicated that she had been asked the questions in previous pregnancies and that although she disclosed, she did not elaborate on her experiences and reported “feeling sad” and like Vani “tries to forget.” Like Vani, Pippa’s distress remained “uninterrupted” by the health service response, despite them both taking the risk to disclose. Both women appeared guarded in their interviews, making interpretation of the impact of these encounters challenging; however, it is clear that the health system response to both women represented lost opportunities for them to receive any of the benefits.

Did Women Agree With Screening?

We asked all participants whether they supported the policy of asking about abuse as part of the assessment. All 32 women in this sample unambiguously agreed that asking women about experiences of abuse at health visits was useful:

I think this is really good. And I really want every hospital to keep this because so many pregnant women go through this. (Kep)

Yeah, it’s a good thing. If somebody needs help, really help, it’s a good thing. (Vani)

If they’re still getting abused then something can get done about it. (Rose)

It is noteworthy that this view was shared by the women who elected not to disclose. Many of their responses indicated that their situations were dynamic and their decisions reflected a single point of time, suggesting the need to provide multiple opportunities for disclosure.

The multiple pathways to impact and associated benefits or lack of benefits are synthesized in Figure 4. This revised model extends the impact derived in the original model with the addition of pathway conditions and the specification of benefits.

Figure 4.

Revised model for the impact of intimate partner violence screening intervention.

Discussion

In this study of 32 women who had experienced recent IPV, the antenatal assessment IPV screening intervention resulted in a positive impact for 24 women, nil positive impact for six women, and negative impact for two women. Women who disclosed, and those who did not, experienced positive impact. Key conditions for positive impact were Care in the way the midwife asked about the abuse and Support and Validation in response to disclosing. The different pathways to positive impact reflect women’s diverse experiences of prior service use, relationship status, contact with statutory agencies, and, critically, the execution of screening. Nil positive impact pathways confirmed the importance of Support and Validation, where its absence was a condition for the two women who reported negative impacts. No factor was sufficient or necessary for positive outcomes from screening; rather, the pathways demonstrated an interplay of the conditions, some unexpected, such as the fact that receipt of a service was not retained in models. All women agreed with the use of routine screening in the antenatal setting, regardless of disclosing or experiencing positive impact, although for some this followed initial apprehension.

The benefits of Naming and Connection from our original model were reaffirmed in this sample, with three new benefits identified: Unburdening, Steps to Safety, and Enabling Informed Care. The original model posited that Connection occurred through the process of Naming, but this was not borne out in the analysis here, nor did any of the other newly defined benefit domains seem to operate as intermediate outcomes. The finding that for six of 32 women routine screening was unremarkable or of minimal impact aligns with the earlier research in which we found eight of 20 women reported similar reactions. Again, as with the earlier study, the majority described valued benefits from the screening intervention. Women’s situations differed, with many women continuing to experience abuse after exiting the relationship, as others have observed (Burke, Mahoney, Gielen, McDonnell, & O’Campo, 2009; Campbell, 1995). This situation and the lack of access to perpetrators’ behavior may help to explain why interventions for IPV struggle to find significant outcomes in RCTs. Recognition of the diversity of women’s needs is not new. The trans-theoretical model (TTM) of change (Prochaska & DiClemente, 1983) has been applied to IPV research in recognition of how women’s readiness and actions may emerge over time (Reisenhofer & Taft, 2013; Schrager, Smith, Heron, & Houry, 2013; Zink, Elder, Jacobson, & Klostermann, 2004). Although TTM has been criticized for presuming a linear progression and that ending the relationship is sufficient to make women safe (Burke et al., 2009), the different benefits we identify, such as Naming and Steps to Safety reflect different needs, stages, and circumstances for women.

The importance of Care echoes earlier research that this condition is central to women’s decisions to disclose their abuse and women’s desire to not be judged and to be respected and listened to by health providers asking about abuse (Bacchus et al., 2016; Koziol-McLain et al., 2008; Snyder, 2016; Spangaro, Koziol-McLain, et al., 2016; Spangaro et al., 2011; Williams, Halstead, Salani, & Koermer, 2016). Support and Validation is included as part of first-line support contained in the WHO’s (2013, p. 16) clinical guidelines for response to IPV. Recognition of Support and Validation as a discrete element which women distinguish within the screening intervention is lent support from research which indicates that midwives are increasingly happy to ask about abuse, but feel ill-equipped to respond effectively when women actually disclose (Eustace, Baird, Saito, & Creedy, 2016). This may help explain the negative experiences of the two women in this study, whose encounters involved Care, but not Support and Validation. Given that both women remained with their partners and were experiencing emotional abuse, it is also possible that the midwives were expressing either bias toward or lack of understanding about these situations. The value of written information for women experiencing IPV has also been identified by others (Burke et al., 2009).

Research suggests adverse effects from routine screening are rare and minimal (Nelson, Bougatsos, & Blazina, 2012; O’Doherty et al., 2015). That said, ineffective health responses do nothing to interrupt the ongoing pattern of violence and women’s experience of entrapment (Family Violence Death Review Committee, 2016; Richardson & Wade, 2010). Failure to provide Continuity of Care contributed to nil positive effects in this study and supported other findings pointing to its importance in disclosing abuse (Eustace et al., 2016; Evans & Feder, 2016).

The five identified benefits also align with the wider literature. As we identified in our original research, the concepts of Naming and Connection link to processes identified by Judith Herman as needed for healing from trauma (Herman, 1992). The value of Naming is further supported by research indicating that many abused women do not recognize their experiences as constituting IPV (Bradbury-Jones, Taylor, Kroll, & Duncan, 2014), which resonates with the TTM second stage—“Contemplation” (Prochaska & DiClemente, 1983). Abused women’s desire for Connection has also been found to foster disclosure (Williams, Gonzalez-Garda, Halstead, Martinez, & Joseph, 2017) and mediate emotional distress (Goodman, Fels Smyth, & Banyard, 2010). Unburdening as a positive outcome also has support from previous literature which found that women find healing from discussing their experiences of abuse (Snyder, 2016). Steps to safety also echoes the TTM third stage—“Action” (Prochaska & DiClemente, 1983). Unifying these different benefits for abused women who experience intervention could be argued to be a strength of this study.

Limitations

While QCA proved a valuable approach for model testing, constraints include the limited number of conditions which can be included in analyses, requiring researchers’ selection of “standout” factors (Befani, 2016). Similarly, the use of binary coding for conditions risks loss of nuance, although our further thematic analysis of benefits and close reading of transcripts helped ameliorate this risk. Challenges are posed in understanding the role of conditions Use of Service and Continuity of Care, which featured for absence, but not presence of outcome. QCA may seem to oversimplify complex reactions, but it is especially valuable in identifying patterns of experience which, in the presence of the same key conditions, lead to an identified outcome. It is possible that women felt constrained about discussing their experiences within the setting of the antenatal clinic, although interviews were conducted in privacy and midwives were not aware of which women participated.

Implications for Policy and Practice

Antenatal screening is widespread in many jurisdictions and is recommended by the WHO for this population. The issues identified here are of relevance across any settings where health providers are aiming for women to have positive experiences from routine screening for IPV. Our synthesized model points to how interventions could be designed with different pathways in mind. For example, this study suggests how women who elect not to disclose and/or who remain with abusive partners can be assisted through screening interventions, particularly through provision of written information and continuity of care. Models of service which prioritize Care in asking, Support and Validation in response to disclosure, and availability of follow-up specialist services will promote positive outcomes for women experiencing abuse. Training for health workers needs to emphasize each of these aspects of screening interventions. Protocols are needed between health and child protection agencies which recognize that IPV and child abuse are intertwined (Family Violence Death Review Committee, 2016), providing clarity regarding grounds for mandatory reporting that address women’s fears.

Implications for Research

There is a need to develop and test interventions for targeted groups drawing on the pathways identified here. Equally, in addition to trials, qualitative studies are needed to ensure that women’s voices inform decision making, including approaches such as realism and QCA, which allow for the underpinning mechanisms of interventions and pathways to outcomes to be identified. Research on whether screening is playing a part in changing social norms is also warranted, given its implementation for over 20 years in many jurisdictions. Finally, research exploring the impact of interventions on the ongoing pain and other negative health consequences women experience as a legacy of abuse is also needed.

Conclusion

QCA provided the opportunity to model IPV routine screening intervention to positive impact, allowing identification of multiple pathways for positive and nil positive impact. Both disclosing and nondisclosing women reported positive impact, with Care in asking combined with Support and Validation by the midwife proving the dominant factors. Domains of benefit identified here point to the subtle but valued ways in which routine screening can support women, with further attention needed by health services to ensure opportunities are not missed to support women who take the risk to disclose IPV.

Footnotes

Acknowledgements

We are indebted to the women who generously shared their stories, the midwives, and social workers at our sites, RAs: Mary-Anne Frail, Jennifer Ruane, and study investigators Julie Swain, Deb Green, Chris Griffiths, Trudy Allende, and Helen Jarman. We thank Mailin Suchting, Jo Holt, Lorna McNamara, and Tamsin Anderson from NSW Health for their support and Charles Ragin and Nicholas Legewie for advice on QCA methods.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Funding for this research was provided by the Australian National Health and Medical Research Council (NHMRC GNT1024908) and the New South Wales Ministry of Health.

ORCID iD

Jo Spangaro

Author Biographies

Jo Spangaro is a senior lecturer in social sciences at the University of New South Wales (UNSW) in Australia, where she teaches in the social work program. Her research addresses health responses to gender-based violence, drawing on her extensive practice and policy experience in this field. Her recent coauthored book addresses therapy with perpetrators, mothers, and children after incest. She is currently working on a multisite study on what sustains health responses to intimate partner violence.

Jane Koziol-McLain is a professor of nursing in the faculty of health and environmental sciences and director of the Centre for Interdisciplinary Trauma Research at Auckland University of Technology, Auckland, New Zealand. Her research focuses on strengthening the health response to family violence.

Alison Rutherford is a sexual health physician and public health physician based at Illawarra Sexual Health Services and a senior lecturer in the Graduate School of Medicine at the University of Wollongong and the School of Public Health and Community Medicine at UNSW. Alison has a particular interest in providing high-quality clinical services to disadvantaged populations within a public health framework, particularly in the areas of contraception; sexually transmitted infections; HIV; sexual assault; lesbian, gay, bisexual, transgender, and intersex (LGBTI) health; and public health aspects of sexual and reproductive health.

Anthony B. Zwi is professor of global health and development at the UNSW, Australia. He heads the health rights and development group (HEARD@UNSW) and has long-standing interest in all aspects of violence, with emphases on collective violence and gender-based violence, and the need for effective policy responses. He is interested in international development and global health policy, and the extent to which they promote equity and social justice and advance human rights. He seeks to promote evidence-informed policies, based upon a diversity of evidence forms and sources, assessed critically and utilized transparently.

References

Bacchus

Bullock

Sharps

Burnett

Schminkey

Buller

A. M.

Campbell

(2016). “Opening the door”: A qualitative interpretive study of women’s experiences of being asked about intimate partner violence and receiving an intervention during perinatal home visits in rural and urban settings in the USA. Journal of Research in Nursing, 21, 345-364. doi:10.1177/1744987116649634

Befani

(2016). Pathways to change: Evaluating development interventions with Qualitative Comparative Analysis (QCA). Surrey, UK: Expertgruppen för biståndsanalys.

Berg-Schlosser

de Meur

(2009). Comparative research design: Case and variable selection. In Rihoux

Ragin

(Eds.), Configurational Comparative Methods: Qualitative Comparative Analysis and Related Techniques (pp. 19-32). Thousand Oaks, CA: Sage.

Berg-Schlosser

de Meur

Rihoux

Ragin

(2009). Qualitative Configurative Analysis as an approach. In Rihoux

Ragin

(Eds.), Configurational comparative methods: Qualitative Comparative Analysis (QCA) and related techniques (pp. 1-18). Thousand Oaks, CA: Sage.

Bradbury-Jones

Taylor

Kroll

Duncan

(2014). Domestic abuse awareness and recognition among primary healthcare professionals and abused women: A qualitative investigation. Journal of Clinical Nursing, 23, 3057-3068.

Burke

Mahoney

Gielen

McDonnell

O’Campo

(2009). Defining appropriate stages of change for intimate partner violence survivors. Violence and Victims, 24, 36-51.

Campbell

(1995). Assessing dangerousness: Violence by sexual offenders, batterers and child abusers. Thousand Oaks, CA: Sage.

Eustace

Baird

Saito

Creedy

(2016). Midwives’ experiences of routine enquiry for intimate partner violence in pregnancy. Women and Birth, 29, 503-510.

Evans

M. A.

Feder

(2016). Help-seeking amongst women survivors of domestic violence: A qualitative study of pathways towards formal and informal support. Health Expectations, 19, 62-73. doi:10.1111/hex.12330

10.

Family Violence Death Review Committee. (2016). Fifth report: 2014 to December 2015. Wellington, NZ: Author.

11.

Feder

Hutson

Ramsay

Taket

(2006). Women exposed to intimate partner violence: Expectations and experiences when they encounter health care professionals: A meta-analysis of qualitative studies. Archives of Internal Medicine, 166, 22-37. doi:10.1001/archinte.166.1.22

12.

Garcia-Moreno

Hegarty

d’Oliveira

Koziol-McLain

Colombini

Feder

(2015). The health systems response to violence against women. Lancet, 385, 1567-1579.

13.

Goodman

Fels Smyth

Banyard

(2010). Beyond the 50-minute hour: Increasing control, choice, and connections in the lives of low-income women. American Journal of Orthopsychiatry, 80, 3-11.

14.

Hackett

Taylor

(2014). Decision making in social work with children and families: The use of experiential and analytical cognitive processes. British Journal of Social Work, 44, 2182-2199.

15.

Hegarty

Gunn

O’Doherty

Taft

Chondros

Feder

, et al. (2010). Women’s evaluation of abuse and violence care in general practice: A cluster randomised controlled trial (weave). BMC Public Health, 10, Article 2. Retrieved from http://www.biomedcentral.com/1471-2458/10/2

16.

Herman

(1992). Trauma and recovery: The aftermath of violence—From domestic abuse to political terror. New York: HarperCollins.

17.

Jordan

Javernick-Will

Amadei

(2014). A qualitative comparative analysis of neighborhood recovery following Hurricane Katrina. International Journal of Disaster Resilience in the Built Environment, 5, 391-412.

18.

Klevens

Kee

Trick

Garcia

Angulo

F. R.

Jones

Sadowski

L. S.

(2012). Effect of screening for partner violence on women’s quality of life: A randomized controlled trial. Journal of the American Medical Association, 308, 681-689.

19.

Klevens

Saltzman

(2009). The controversy on screening for intimate partner violence: A question of semantics? Journal of Women’s Health, 18, 143-145.

20.

Kothari

Rohs

Davidson

Kothari

Klein

Koestner

, et al. (2015). Emergency department visits and injury hospitalizations for female and male victims and perpetrators of intimate partner violence. Advances in Emergency Medicine, 2015, Article 502703. doi:10.1155/2015/502703

21.

Koziol-McLain

Garrett

Fanslow

Hassall

Dobbs

Henare-Toka

Lovell

(2010). A randomized controlled trial of a brief emergency department intimate partner violence screening intervention. Annals of Emergency Medicine, 56, 413-423.

22.

Koziol-McLain

Giddings

Rameka

Fyfe

(2008). Intimate partner violence screening and brief intervention: Experiences of women in two New Zealand health care settings. Journal of Midwifery & Women’s Health, 53, 504-510.

23.

MacMillan

H. L.

Wathen

C. N.

Jamieson

Boyle

McNutt

L. A.

Worster

, et al. (2006). Approaches to screening for intimate partner violence in health care settings: A randomized trial. Journal of the American Medical Association, 296, 530-536.

24.

MacMillan

H. L.

Wathen

C. N.

Jamieson

Boyle

Shannon

Ford-Gilboe

, et al. (2009). Screening for intimate partner violence in health care settings: A randomized trial. Journal of the American Medical Association, 302, 493-501. doi:10.1001/jama.2009.1089

25.

Nelson

Bougatsos

Blazina

(2012). Screening women for intimate partner violence: A systematic review to update the 2004 U.S. preventive service task force recommendation. Annals of Internal Medicine, 156, 796-808.

26.

NSW Department of Health. (2006). Policy and procedures for identifying and responding to domestic violence. North Sydney, Australia: NSW Health.

27.

NSW Ministry of Health. (2016). Domestic violence routine screening November 2015 snapshot 13. North Sydney, Australia: Author. Retrieved from https://www.health.nsw.gov.au/parvan/DV/Pages/dvrs.aspx

28.

O’Doherty

Hegarty

Ramsay

Davidson

Feder

Taft

(2015). Screening women for intimate partner violence in healthcare settings. Cochrane Database of Systematic Reviews, 7, Article CD007007.

29.

Prochaska

DiClemente

(1983). Stages and processes of self-change of smoking: Toward an integrative model of change. Journal of Consulting and Clinical Psychology, 51, 390-395.

30.

Ragin

(2008). Users guide to fuzzy-set/qualitative comparative analysis. Tucson: University of Arizona. Retrieved from http://www.socsci.uci.edu/~cragin/fsQCA/download/fsQCAManual.pdf

31.

Ragin

(2014). The comparative method: Moving beyond qualitative and quantitative strategies (2nd ed.). Oakland, CA: University of California Press.

32.

Ragin

Davey

(2014). fs/QCA (Version 2.5). Irvine: University of California. Retrieved from http://www.compasss.org/software.htm#fsQCA4/5/2015

33.

Reisenhofer

Taft

(2013). Women’s journey to safety—The Transtheoretical Model in clinical practice when working with women experiencing intimate partner violence: A scientific review and clinical guidance. Patient Education & Counseling, 93, 536-548. doi:10.1016/j.pec.2013.08.004

34.

Richardson

Wade

(2010). Islands of safety: Restoring dignity in violence-prevention work with Indigenous families. First Peoples Child & Family Review, 5, 137-145.

35.

Rihoux

de Meur

(2009). Crisp-set qualitative comparative analysis. In Rihoux

Ragin

(Eds.), Configurational comparative methods: Qualitative Comparative Analysis (QCA) and related techniques (pp. 33-68). Thousand Oaks, CA: Sage.

36.

Rihoux

Ragin

(Eds.) (2009). Configurational comparative methods: Qualitative Comparative Analysis (QCA) and related techniques. Thousand Oaks, CA: Sage.

37.

Ritchie

Spencer

(1994). Qualitative data analysis for applied policy research. In Bryman

Burgess

R. G.

(Eds.), Analysing qualitative data (pp. 173-194). London, England: Routledge.

38.

Rivas

Ramsay

Sadowski

Davidson

Dunne

Eldridge

, et al. (2015). Advocacy interventions to reduce or eliminate violence and promote the physical and psychosocial well-being of women who experience intimate partner abuse. Cochrane Database of Systematic Reviews, 12, CD005043. doi:10.1002/14651858.CD005043

39.

Schrager

J. D.

Smith

L. S.

Heron

S. L.

Houry

(2013). Does stage of change predict improved intimate partner violence outcomes following an emergency department intervention? Academic Emergency Medicine, 20, 169-177. doi:10.1111/acem.12081

40.

Snyder

(2016). Women’s experience of being interviewed about abuse: a qualitative systematic review. Journal of Psychiatric and Mental Health Nursing, 23(9-10), 605-613. doi: 10.1111/jpm.12353

41.

Spangaro

Herring

Frail

Koziol-McLain

Rutherford

Zwi

(2016). “They aren’t really black fellas but they are easy to talk to”: Factors which influence Australian Aboriginal women’s decision to disclose intimate partner violence during pregnancy. Midwifery, 41, 79-88.

42.

Spangaro

Herring

Koziol-McLain

Rutherford

Zwi

(2019). “Yarn about it”: Aboriginal Australian women’s perceptions of the impact of routine inquiry for intimate partner violence. Culture Health and Sexuality, 21(7), 789-806. doi: 10.1080/13691058.2018.1519117

43.

Spangaro

Koziol-McLain

Zwi

Rutherford

Frail

Ruane

(2016). Deciding to tell: Qualitative configurational analysis of decisions to disclose experience of intimate partner violence in antenatal care. Social Science & Medicine, 154, 45-53.

44.

Spangaro

Zwi

A. B.

Poulos

(2009). The elusive search for definitive evidence on routine screening for intimate partner violence. Trauma, Violence, & Abuse, 10, 55-68. doi:10.1177/1524838008327261

45.

Spangaro

Zwi

A. B.

Poulos

(2011). “Persist. Persist.”: A qualitative study of women’s decisions to disclose and their perceptions of the impact of routine screening for intimate partner violence. Psychology of Violence, 1, 150-162.

46.

Sprague

Goslings

Hogentoren

de Milliano

Simunovic

Madden

Bhandar

(2014). Prevalence of intimate partner violence across medical and surgical health care settings: A systematic review. Violence Against Women, 20, 118-136.

47.

Stark

(2007). Coercive control: How men entrap women in personal life. New York: Oxford University Press.

48.

Watts

Heise

Ellsberg

Garcia Moreno

(2001). Putting women first: Ethical and safety recommendations for research on domestic violence against women. Geneva, Switzerland: World Health Organization.

49.

Wilbur

Noel

Couri

(2013). Is screening women for intimate partner violence in the emergency department effective? Annals of Emergency Medicine, 62, 609-611. doi:10.1016/j.annemergmed.2013.06.012

50.

Williams

Gonzalez-Garda

Halstead

Martinez

Joseph

(2017). Disclosing gender-based violence during health care visits: A patient-centered approach. Journal of Interpersonal Violence. doi:10.1177/0886260517720733

51.

Williams

Halstead

Salani

Koermer

(2016). Intimate partner violence screening and response: Policies and procedures across health care facilities. Women’s Health Issues, 26, 377-383. doi:10.1016/j.whi.2016.03.006

52.

Willis

(2011). Trends and Issues in Crime and Criminal Justice. Non-disclosure of violence in Australian Indigenous communities. Canberra: Australian Institute of Criminology.

53.

World Health Organization. (2013). Responding to intimate partner violence and sexual violence against women: WHO clinical and policy guidelines. Geneva, Switzerland: Author.

54.

Zakrison

Ruiz

Gelbard

Cline

Turay

Luo-Owen

, et al. (2017). Universal screening for intimate partner and sexual violence in trauma patients: An EAST multicenter trial. Journal of Trauma and Acute Care Surgery, 83, 105-110. doi:10.1097/TA.0000000000001495

55.

Zink

Elder

Jacobson

Klostermann

(2004). Medical management of intimate partner violence considering the stages of change: Precontemplation and contemplation. Annals of Family Medicine, 2, 231-239.