Barriers and facilitators of disclosing domestic violence to the healthcare service: A systematic review of qualitative research

Abstract Domestic violence victims are in frequent contact with the healthcare service yet rarely disclose. Therefore, it is critical to understand victims' experiences and perceptions regarding disclosure in healthcare settings. The goal of this review is to provide an updated synthesis of qualitative research identifying barriers and facilitators, advice, and positive and negative outcomes of adult victims' disclosure of domestic violence to healthcare professionals (HCPs). A systematic search of PsychINFO, CINAHL and Web of Science was conducted in January 2018. Thirty‐four eligible studies were identified, including 783 domestic violence victims (781 females). Formal quality assessment indicated variable study quality. Barriers of disclosure included negative HCPs attitudes, victims' perceptions of safety and concerns about the consequences of disclosing. Facilitators of disclosing included a positive relationship with the HCP, HCPs directly asking victims about abuse, and HCPs ensuring that the environment is safe and disclosure is confidential. Victims advised increased awareness of HCPs reactions to disclosure and avoiding mirroring their perpetrators minimization. HCPs were encouraged to engage in direct questioning and maintain a supportive and secure environment. Positive and negative outcomes of abuse were identified, such as being able to leave the abuser or, on the other hand, the victims' situation not changing. Our results indicate that barriers for disclosure of domestic violence in healthcare settings persist despite the widespread implementation of policies and guidelines to counter them. Based on these findings, we provide recommendations for clinical practice and future research to help improve disclosure in healthcare settings.

vices identification and responses to disclosures from victims of domestic violence have been introduced (Department of Health, 2000;World Health Organisation, 2013b). Direct questioning is an effective strategy facilitating domestic violence victims' disclosures (Cann et al., 2001;Howard et al., 2010). Disclosure can help victims to become free from violence within 6 weeks in situations when they are given appropriate care following disclosure (Krasnoff & Moscati, 2002).
Yet, despite healthcare providers (HCPs) significant advantage in accessing this hard to reach population and the potential benefits of disclosure, HCPs have not always been effective in identifying victims of domestic violence (Campbell et al., 2002;Chapman & Monk, 2015;Feder et al., 2011). Only 10%-50% of the domestic violence cases are detected in healthcare services (Feder et al., 2006;Gremillion, & Kanof, 1996). A barrier to disclosure may be that HCPs do not feel capable or comfortable discussing domestic abuse (Taylor et al., 2013). Additionally, HCPs have reported that a lack of time, privacy, training, resources and knowledge on how to ask about domestic violence have prevented them from enquiring about abuse (Beynon et al., 2012;Sundborg et al., 2012). For example, doctors and nurses report receiving little or no training in responding to domestic violence (Rimmer, 2017;Taft et al., 2004).
Perhaps unsurprisingly, victims have reported that HCPs have been inappropriate, inadequate and unhelpful in responses to disclosures of abuse (Pratt-Erickson et al., 2014;Trevillion et al., 2011Trevillion et al., , 2014. To improve responses from HCPs, it is also important to understand how victims perceive and experience disclosure in healthcare settings. More than a decade ago, a systematic review of qualitative studies by Robinson and Spilsbury (2008) on disclosing domestic abuse to HCPs first summarized findings on this topic. They charted what barriers and facilitators of disclosure domestic violence victims reported. They found that victims wanted the topic of abuse to be routinely raised by HCPs to make it easier to disclose, yet also had concerns about disclosing. For example, victims felt that just one consultation with a professional was not enough to build the trust needed to disclose and that the brevity of appointments with their HCP limited opportunities for disclosure. Victims also reported that a lack of privacy at their healthcare setting prevented them from disclosing. Other major barriers were victims' fears that they would lose their children or that the abuse would escalate. Robinson and Spilsbury's (2008) review focussed on the barriers and facilitators that victims experience and perceive when disclosing abuse to the health service, yet they did not look specifically at victims' advice provided to HCPs or their reported outcomes of disclosure. It appears important to also summarize such advice as victims themselves could be viewed as "experts" of their own experiences enabling them to bring unique insight and knowledge (Reid et al., 2005) which may improve future victims' experiences of disclosing abuse.
Additionally, it appears critical to investigate victims' reported outcomes of disclosure. For example, a positive outcome might be a victim experiencing direct change after disclosing such as them leaving their partner or filing a police report (Liebschutz et al., 2008), whereas a negative outcome of disclosure could be the victim feeling responsible for the abuse and becoming convinced that nothing will change (Damra et al., 2015). Charting such outcomes could provide further insight into the potential benefits or drawbacks of disclosing which in turn could help motivate HPCs to act in line with established guidelines and regulations to facilitate disclosure. The current review therefore sought to update and extend Robinson and Spilsbury's (2008) review, to provide a more comprehensive, up-todate overview of qualitative research of victims' views and experiences related to disclosure.

What is known about this topic?
• Domestic violence victims are in frequent contact with healthcare professionals, yet rarely disclose.
• Disclosure can help victims become free from violence.
• International guidelines exist to facilitate disclosure in healthcare settings What this paper adds?
• Our up-to-date synthesis of qualitative research identified a variety of barriers and facilitators of disclosure, victim advice on disclosure, and outcomes of disclosure in healthcare settings.
• Barriers and facilitators identified in previous studies persisted, yet we also identified new barriers and facilitators.
• Healthcare professionals are encouraged to engage in direct questioning and maintain a supportive and safe environment. Positive outcomes of disclosure included feeling validated and leaving the abuser. Negative outcomes included the stigmatization and continued abuse.

| Aim of the review
While Robinson and Spilsbury (2008) provided an important starting point to better understand barriers and facilitators of disclosure, their review was limited in size (including 10 papers), scope (focusing only on barriers and facilitators and including only studies from English-speaking samples) and rigour (study quality was not systematically assessed). To overcome these limitations, we conducted an updated systematic review of qualitative studies that investigated experiences and perceptions of domestic violence victims on disclosure in healthcare settings. This review includes a decade of new research on barriers and facilitators, expands the scope of the review to include victims' advice and outcomes of disclosure, and weighs the evidence by providing a systematic assessment of study quality.
Specifically, the review aimed to address the following questions:

| ME THOD
This review was conducted in agreement with the guidelines and criteria for systematic reviews reporting set out by the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement (Moher et al., 2009). This review was not pre-registered as the preparation for this review started before PRISMA was widely applied.

| Search strategy
A comprehensive search of three electronic databases, PsychINFO, CINAHL and Web of Science, was conducted on the 30 January 2018 (see Appendix S1). The keywords used in the search strategy were drawn from the previous systematic review conducted by Robinson and Spilsbury (2008). Additional search terms were added by checking the keywords used in articles identified in preliminary searches. A Boolean approach was used, and the following search terms were chosen after scoping the literature: ["Domestic* violence*" OR "battered female*" OR "intimate partner violence" OR "partner abuse*" OR "domestic abuse*" OR "battered men*" OR "battered male**" OR "battered women*" OR "victim*" OR "spouse abuse*" OR "survivor*" OR "female survivor*" OR "male survivor*" OR "intimate partner violence survivor*"] AND ["Self-Disclosure" OR "Disclosure*" OR "help seeking"] AND ["Health care service*"OR "Health care professional*" OR "Health Care Clinician*" OR "Health setting*" OR "Health Care Provider*" OR "nurse*" OR "doctor*" OR "primary care setting*" OR "antenatal service*" OR "mental health service*"].

| Inclusion/Exclusion criteria
As the decision regarding which study designs to include in the review should be dictated by the review question (Nutbeam & Harris, 2004;Petticrew & Roberts, 2003), we included only qualitative studies as these would fully capture the victims' experiences and views of disclosing to the healthcare service. We also considered it important to synthesise qualitative evidence as this has been found to make a positive contribution to the knowledge available to international organisations, such as the World Health Organisation, when developing recommendations on public health topics (Metin Gülmezoglu et al., 2013).
Studies were further included if the study sample consisted of adult (16 years or older) domestic violence victims, who had experienced partner abuse. Thus, studies including victims of violence from other family members than the partner were excluded.
Furthermore, the study should describe victims' experiences of disclosure/interactions within healthcare services (i.e., with professionals with health-related qualifications, e.g., doctors, nurses, midwives). To safeguard study quality, readability and interpretability, we only included papers that were published in peer-reviewed English language scientific journals. To capture most recent developments on barriers and facilitators of disclosure, we included papers published between 1996 and January 2018. Lastly, low quality studies (as determined by an adapted Critical Appraisal Skills Programme Checklist, CASP -see Section 3.4) were excluded.

| Study selection
All references were exported to Endnote Web. Our searches identified 647 papers, which was reduced to 489 after removal of duplicates. Titles and abstracts, as well as full text articles, were screened independently against inclusion and exclusion criteria by two reviewers. Disagreements were resolved through discussion until consensus was reached. Forty-two papers underwent full text review, of which 32 studies met inclusion criteria. Additionally, five articles were identified by screening the reference lists of included studies.
Quality assessment led to the exclusion of three low quality papers, reducing the total number of studies to 34. For a PRISMA flowchart see Figure 1.

| Quality appraisal
The quality assessment was conducted with an adapted version of the CASP checklist for qualitative studies (Critical Appraisal Skills Programme, 2017). The quality of the studies was assessed on the domains research design, sampling and selection bias, attrition bias, performance bias, data collection, data analysis, and ethical issues (see Appendix S2 for the adapted checklist). Fourteen questions were answered to assess study quality in these domains and were answered positively with yes (+), negatively with no (−), or with not reported (N.R). A total score, number of unclear items, and a quality assessment score was given for each included article. The quality was rated on a rating scale from A -best (no risk of bias or only risk of bias in one of the assessed domains), B -good (risk of bias in two domains), C -sufficient (risk of bias in three domains), to D -insufficient (risk of bias in four or more domains). If the quality assessment score was D the study was excluded (see Appendix S3 for the methodological quality of included studies). To ensure reliability and uniformity of quality assessment, it was conducted independently by two evaluators. Any discrepancies in agreement were discussed and resolved and consensus was reached on the final quality assessment ratings by both reviewers.

| Data extraction
Data were extracted on sample characteristics (e.g., age, gender), recruitment strategy and location, qualitative research methods and analysis, and findings of relevance to our research questions (barriers and facilitators, advice, and outcomes of disclosure), independently by two researchers. Differences in extracted information were discussed until consensus was reached.

| Study characteristics
For a summary of findings from the included studies see Table 1.

Barriers:
Fear of retaliation. Fear that children would be taken away. Feelings of shame, humiliation and embarrassment. Negative attitudes of the HCP. HCPs being too busy and the emergency room not being the appropriate place to discuss such topics.

Outcomes:
Seven of the women reported positive outcomes, they felt that the HCP was helpful and listened to their needs. The HCP did not pressure them to leave their abuser, but gave them support and information.
Nine women reported negative experiences, they felt judged by the HCP. Other women felt ignored or stigmatized and some women were even made fun of.

Q1 and 3 C
Hathaway et al.

Advice:
HCPs should be trained on how to react and deal with a DV disclosure. It is important that HCPs validate and support women who disclose DV to them. HCP should try not to dismiss disclosures as it could mirror perpetrators behaviours of minimization.

Outcomes:
Women who had positive responses (i.e., appropriate support and advice offered) after disclosure were able to leave violent relationships more often than women who had negative responses (i.e., whose disclosures were dismissed) when disclosing.

Barriers:
Fear that confidentiality will be broken and that they will lose their children. Since domestic violence is not viewed as a health topic, not all women felt it could be discussed with their HCP. No practical support or protection after disclosing. Language barrier prevented a victim from disclosing (need for interpreters).

Facilitators:
A good relationship with the HCP and privacy. The women who disclosed all did so during a home visit.

Barriers:
Fear, including fear of social services; fear that they will not be believed; fear that disclosure will lead to more violence; fear to disrupt the family and fear of the consequences for one's immigration status. Self-blame, shame and embarrassment. Social isolation and lack of a support network.

Facilitators:
A good and supportive relationship with the HCP.

| Study quality
Three of the studies were of insufficient quality and thus excluded.
Sixteen of the studies included in our review were rated as A -best (no risk of bias or only risk of bias in one of the assessed domains), 12 were rated as B -good (risk of bias in two domains) and six were rated as C -sufficient (risk of bias in three domains). The quality criteria that were most often judged as low was that of attrition bias (in all 34 included studies) and performance bias (in 16 included studies). Study quality was not explicitly considered when interpreting the findings, as the majority of studies were of good quality and variability in study quality was limited.

| Main findings
Themes for each article were identified and grouped under three main headings: barriers and facilitators of disclosure, advice to improve disclosure and outcomes of disclosure.

Barriers:
Feelings of shame, embarrassment, fear for partner and concern for their children. Feeling that the HCP does not understand their situation, does not take it seriously or in some cases that the HCP blames them for their situation. Negative attitudes of and relationship with the HCP. The HCP seems uncaring, unsympathetic, rushed or unwilling to discuss abuse. Lack of resources/support system. Advice: HCPs should express compassion, ensure advocates are available, make the women feel safe, provide them with options, listen to them and discuss disclosure privately. Follow-up on the patient. Have a trained counsellor and advocate for victims available.

Q1 and 2 B
Zink et al.

Barriers:
Feelings of shame, commitment to their abuser and not realizing they were in an abusive relationship. Victims also felt that their upbringing had taught them to stay with their husband and that they should not challenge him. Lack of time and privacy. Advice: Listen to needs of patients and having an empathic response. Not judging the women for not leaving or pressuring them into leaving. Raising the subject of DV, so women know they can talk about it Outcomes: Several of the women had negative experiences after disclosing. Providers ignored their disclosure or were uncomfortable with discussing it. This caused these women not receiving the validation or assistance they were seeking. There was also a lack of empathy because the women would not leave their husbands. In contrast over half the women felt positive about their disclosure. Their HCPs were helpful, listened and gave referrals for support, they validated the women and in some cases, this lead to the women leaving their abuser Q1, 2 and 3 B Abbreviations: advice; DV, domestic violence; facilitators and barriers; GP, general practitioner; HCP, healthcare professional; IPV, intimate partner violence; outcomes; Q1, question 1; Q2, question 2; Q3, question 3.

| Barriers and facilitators of disclosure
Of the 34 articles included in this review, 28 discussed barriers and/ or facilitators for disclosure. Most of the articles (27) discussed barriers, and 11 (additionally) discussed facilitators.
Thirdly, victims communicated that fear of their abuser further prevented them from disclosing. For instance, they were fearful of what would happen if their abuser found out about their disclosure and they were worried about whether the healthcare service would be able to protect them, especially as in many cases their abuser made threats prior to them disclosing (Hathaway et al., 2002;Hegarty & Taft, 2001;Kelley, 2006;Lutenbacher et al., 2003;McCauley et al., 1998;Narula et al., 2012;Othman et al., 2014;Peckover, 2003;Rodriguez et al., 1996;Rose et al., 2011;Spangaro et al., 2011;Yam, 2000). Relatedly, three studies found that victims feared that confidentiality would be broken by the health service personnel (Gerbert et al., 1999;Peckover, 2003;Wallin et al., 2003) and they were also concerned about the lack of privacy at the healthcare service (Bacchus et al., 2003;Bates et al., 2001;Damra et al., 2015;McCauley et al., 1998;Reisenhofer & Siebold, 2013;Zink et al., 2004).
A lack of a positive relationship with the HCP was also viewed as a barrier to disclosure (Bacchus et al., 2003;Bates et al., 2001;Damra et al., 2015;Hathaway et al., 2002;Kelly, 2006;Narula et al., 2012;Nicolaidis et al., 2008;Rishal et al., 2016;Spangaro, Herring, et al., 2016;Wallin et al., 2018;Yam, 2000). This included not trusting in the HCP, lack of continuity in the relationship or limited time with the HCP, and not expecting that the HCP would be empathetic.
A further barrier was the fact that many victims thought that healthcare services were not the appropriate place to discuss abuse, as they perceived that it was not a health issue that you discuss with a HCP (Gerbert et al., 1997;Narula et al., 2012;Othman et al., 2014;Peckover, 2003). Additionally, victims were also not always aware of their rights in terms of what choices or support they could gain by telling the HCP about abuse (Hathaway et al., 2002;Kelly, 2006;Othman et al., 2014).
Additional factors facilitating victims' disclosures were perceiving the HCP as knowledgeable and capable to handle the abuse (Battaglia et al., 2003;Hathaway et al., 2002), and the gender of the HCP; a woman facilitated disclosure among female victims (Bates et al., 2001). Victims further noted that feeling that they had a choice over what to disclose and how much, helped them to disclose (Spangaro, Herring, et al., 2016;Spangaro et al., 2011). Finally, having leaflets and posters on domestic violence visible and available in waiting rooms further aided victims' disclosures (Bates et al., 2001).

| Advice to improve disclosure
Of the 34 studies included in the review, 16 contained advice from victims on how the healthcare service may improve responding to victims' disclosures.
A commonly given advice was to make victims aware of the resources and options available to disclose in the healthcare service.
It was further advised that HCPs put in an effort to make the environment safe, private and confidential for disclosure (Chang, Decker, et al., 2005;Rishal et al., 2016;Rodriguez et al., 1996;Yam, 2000). HCPs were also advised to be aware of how their attitude could influence the choice of a victim to disclose. Victims advised that a caring, non-judgemental and supportive attitude is what would facilitate disclosure and they stressed the importance of respecting the victim and showing compassion (Keeling & Fisher, 2015;McCauley et al., 1998;Nicoladis et al., 2008;Rishal et al., 2016;Rodriguez et al., 1996;Yam, 2000;Zink et al., 2004).
Female victims also communicated finding it easier to talk about abuse to another woman (Damra et al., 2015) and so they advised that victims should be provided with an option to meet with a female HCP. Furthermore, it was advised that HCPs provide referrals and continued support (Narula et al., 2012;Rodriguez et al., 1996;Yam, 2000) and that they are specially trained to deal with domestic violence victims disclosures (Bates et al., 2001). It was also advised to make a mental health counsellor and a legal counsellor available (Chang, Cluss, et al., 2005;Yam, 2000). Lastly, victims recommended not to pressure patients into leaving their abuser (Hathaway et al., 2002;Zink et al., 2004).

| Outcomes of disclosure
Nine of the 34 included articles discussed what victims experienced after their disclosure of domestic abuse to a HCP. The studies discussed both positive and negative outcomes of disclosure of domestic violence.
The latter steps appear particularly likely to be undertaken when initial responses to disclosure were a positive experience.
However, some of the studies from our review also reported negative outcomes of disclosure. For example, victims noted no change or improvement in their situations after disclosing (Damra et al., 2015;Wong et al., 2008) and instead feeling blamed, stigmatized or ignored by HCPs (Damra et al., 2015;Gerbert et al., 1997;Zink et al., 2004). After disclosing, some victims also reported feeling helpless and fearful (Wong et al., 2008) noting that they felt even more endangered after disclosing abuse (Liebschutz et al., 2008).

| D ISCUSS I ON
The aim of this systematic review was to synthesize qualitative research on the barriers and facilitators of disclosure that victims of domestic violence experience in healthcare settings. We also synthesized the advice that victims provide on the ways in which healthcare services can increase victims' disclosures of domestic abuse, and what outcomes they perceived after disclosing domestic abuse to a HCP.
Generally speaking, this review has demonstrated a continued scientific interest in domestic violence victims' experiences of disclosure in healthcare settings. Over the past 12 years, more than twice as many qualitative papers were written about this topic than in the 10 years preceding it. A majority of studies focus on barriers to disclosure, but facilitators of disclosure, victims' advice to HCPs, and outcomes of disclosure are increasingly investigated as well.
Barriers to disclosure within this review can be divided into two parts. Firstly, there are barriers related to opinions and feelings of the victims. These barriers consisted of victims' fear of their children being taken away by child protection services, their abuser, and being negatively judged by the HCP and/or their social environment.
When comparing present results against Robinson and Spilbury's (2008)  identified. This included victims fearing not being believed by HCPs (Narula et al., 2012;Rose et al., 2011;Salmon et al., 2015), being concerned about re-living the trauma when disclosing to a HCP , having perception that the HCP is incapable to respond to disclosure (Damra et al., 2015), and a lack of eye contact with clients due to the use of computer screens  World Health Organisation, 2013b) and repeated recommendations by researchers on how HCPs can facilitate disclosure of abuse Nyame et al., 2013), our review illustrates both the variety of barriers to disclosure and the continuity of such barriers over time and potentially across cultures.
Similarly, identified facilitators show some consistency across time. For example, both Robinson and Spilsbury's (2008) review and our review found that a positive and trusting relationship with the HCP, safety, the HCP being knowledgeable and ensuring that leaflets and posters were available within the healthcare setting facilitated disclosure. However, our review also identified new facilitators of disclosure such as victims having autonomy over what to disclose and how much (Spangaro, Herring, et al., 2016;Spangaro et al., 2011) and them being questioned directly about abuse (Hathaway et al., 2002;Spangaro, Herring, et al., 2016;Spangaro et al., 2011;Wong et al., 2008). The findings from both our review therefore highlight that disclosure can be improved for victims. We recommended that HCPs be educated upon these ways with training, especially as training can help HCPs identify and respond better to victims of domestic abuse (Beynon et al., 2012).
Relatedly, when comparing results from our review with Robinson and Spilsbury's (2008) review, some consistency on advice provided by domestic violence victims to HCPs is apparent. While the prior review did not specifically set out to synthesize advice, it did show that when victims provide advice, this was in many respects similar as the advice identified in our review. For instance, many of the included studies in this review recommended that HCPs ensure that posters and pamphlets are made available (Bates et al., 2001;Chang, Cluss, et al., 2005;Chang, Decker, et al., 2005;Dienemann et al., 2005;Kelly, 2006;Rishal et al., 2016), that the environment is kept private and confidential to help victims to feel safe (Chang, Cluss, et al., 2005;Rishal et al., 2016;Yam, 2000;Zink et al., 2004), that victims disclosures should not be judged, that victims should not be pressured into leaving (Hathaway et al., 2002;Rishal et al., 2016;Zink et al., 2004), and that victims be provided with counselling and continued support after disclosing (Chang, Decker, et al., 2005;Narula et al., 2012;Rishal et al., 2016). However, our review also uniquely identified new advice such as that direct questioning should be carried out by HCPs (Damra et al., 2015;Kelly, 2006;Spangaro, Herring, et al., 2016). It should be noted though that screening has not always been supported in previous studies as it may negatively impact on women's mental health (Klevens et al., 2012). Other new advice included providing training to HCPs on how to react and deal with a domestic violence disclosure and avoiding minimizing victims' experiences of domestic violence when disclosing, as this could mirror perpetrators minimization of abuse (Keeling & Fisher, 2015).
With respect to the outcomes of disclosure, which was not covered by Robinson and Spilsbury (2008), a nuanced picture emerges.
On the one hand, disclosure can help victims to feel validated and supported (Gerbert et al., 1997;Hathaway et al., 2002;Zink et al., 2004), which can be experienced as a turning point (Gerbert et al., 1999;Spangaro et al., 2011) resulting in victims deciding to leave their abuser (Liebschutz et al., 2008), or becoming aware of what support is available (Keeling & Fisher, 2015;Wong et al., 2008). Such findings are generally in line with quantitative research, which has for example demonstrated that approximately half of domestic violence victims can become free from violence within 6 weeks after disclosure if provided with appropriate care (Krasnoff & Moscati, 2002). However, on the other hand, potential negative consequences of disclosure should not be disregarded. For example, negative responses from HCPs can lead to victims feeling ignored, stigmatized (Gerbert et al., 1997;Zink et al., 2004), and blamed (Damra et al., 2015), resulting in them finding it more difficult to leave. Our review also highlighted how some victims could feel even more endangered (Liebschutz et al., 2008) and helpless to change their situations (Wong et al., 2008).

| Implication for policy and practice
Based on the above, we conclude that training is important and should be provided to HCPs to help them to understand the complexities of the disclosure process and the variety of potential barriers and facilitators of disclosure that may be experienced. For example, research shows that HCPs would rather not ask about domestic violence because it makes them uncomfortable or afraid of offending their patients (Hamberger et al., 1998). However, several studies in our review showed that victims of domestic abuse would prefer being asked directly about abuse (e.g., Damra et al., 2015;Kelly, 2006 (Fulfer et al., 2007), and then gradually they could ask more direct questions such as, 'Are you experiencing abuse emotional, physical or sexual?' Training has been shown to be more effective if professionals are given the opportunity to observe and model good practice. For example, experiential and interactive training has been found to be effective in previous studies with HCPs (Haney et al., 2003;Zaher et al., 2014). Haney et al. (2003) demonstrated that HCPs confidence in asking about and responding to domestic violence markedly improved after completing an interactive training.
Additionally, healthcare services should take into account barriers unrelated to HCPs training and abilities. For instance, a lack of information about domestic violence, a lack of privacy or a lack of opportunity to disclose to a female HCP, all act as barriers to disclosure, yet could be targeted by implementing structural changes in healthcare facilities (Bacchus et al., 2003;Bates et al., 2001;McCauley et al., 1998;Reisenhofer & Seibold, 2013;Zink et al., 2004). For example, creating private areas for contact with female HCPs in Accident and Emergency hospitals or maternity clinics could directly reduce a barrier to disclosure. This might also help reduce the emotional barrier of feelings of shame and embarrassment. That is, if fewer people interact with a victim in a private environment, this could make her feel more comfortable to openly discuss her experiences.
Additionally, based on the findings from our review, we concluded that victims may find it difficult to disclose due to fears that their children will be removed. Victims' anxiety may be warranted as guidelines have highlighted that general practitioners do not always know what to do when presented with child safeguarding cases of domestic violence (General Medical Council, 2012). Thus, it is recommended that training is provided to HCP's and that they are reminded to be sensitive towards victims, making every effort not to blame them for difficulties they experience in protecting their children due to domestic violence (Lapierre, 2008;Radford & Hester, 2006). Instead HCP's should work closely with child protective services to ensure that victims are fully supported to be able to care and protect their children. Research by Mullender et al. (2002) also showed that children (aged 8-17) who were living in homes with domestic violence, wished to be treated as agentic and be involved in the decision-making process on solutions to this problem, including whether they want to remain in their mother's care. These measures, combined with the provision of information on how child safeguarding is handled by healthcare professionals (e.g., in information brochures on domestic violence) could reduce victims fears about losing their children after disclosing.

Finally, after acknowledging victims' recommendations for what
can improve disclosure, we also must not forget that disclosure does not always lead to positive consequences and may result in more harm if not responded to correctly (Liebschutz et al., 2008). Victims in one study advised that HCPs should be trained specifically on how to respond and deal with a disclosure (Keeling & Fisher, 2015). We need to ensure that we optimize the possibility of positive outcomes for victims when disclosing, especially since disclosure can lead to victims becoming free from violence if they are provided with appropriate care (Krasnoff & Moscati, 2002).
Appropriate care based on the qualitative findings from this review from the perspective of the victims would consist of the HCP responding in an empathic (e.g., 'I am so sorry that this is happening to you'), validating (e.g., 'you are correct what you are experiencing is abuse'), non-judgemental (e.g., 'You do not deserve this and none of this is your fault') and non-pressurizing way (e.g., 'feel free to discuss with me when you feel ready'), within a safe and private environment (in a closed soundproof room without their partner present).
Such a response from The HCP will allow the victim to (begin) establishing trust with the HCP and form a better relationship. To increase the odds of positive outcomes of disclosure, the HCP should have up to date information on where to refer the victim for appropriate support (e.g., women's shelters, legal advocacy), or care (e.g., counselling) so that the victim can be made aware of all their options.
While not all victims will be at imminent risk of violence, the HCP should assess the victim's risk of violence, which would allow them to create a safety plan with the victim if needed, e.g., advising the victim to store money/documents etc so they can escape urgently if needed. HCPs should ideally also be trained to use tools, which could help them to assess the victim's risk, i.e., screening measures such as the woman abuse screening tool (Basile et al., 2007). Finally, after disclosure, the HCP should end their consultation by acknowledging the victim's courage to disclose abuse ('I am very proud of you, it must not be easy to open up about your experiences') and provide a follow-up appointment if needed.

| Limitations and future research
Multiple considerations should be taken into account when interpreting our findings. Some limitations relate to methodological choices. First, we only synthesised qualitative studies, as this would provide us with rich data, which would help us to understand victims' experiences and perspectives through their own voices. Due to the subjective nature of qualitative research various biases may have shaped results from our review. For example, social desirability may have influenced answers in both the interview or focus group studies, as participants may have adapted their answers due to wanting to be accepted or liked by the researchers or group members. A further limitation of this review was that it did not consider non-English peer reviewed articles, or grey literature (dissertations, master theses), to safeguard interpretability of results. Future reviewers focusing on these topics could aim to be more comprehensive in their selection of manuscripts. A further limitation of our review was that we grouped findings on our main research questions together for all types of HCPs. While we are confident that the conclusions from our review are broadly applicable, some barriers and facilitators may be unique to specific settings or HCPs. For instance, unique barriers to victim disclosure exist within mental healthcare (Rose et al., 2011;Trevillion et al., 2016), such as concerns that mental illness is viewed as the main cause of abuse, rather than (other) personal or social factors (Du

Mont & Forte, 2014).
Other limitations pertain to the acquired data. One of the limitations of this study is the lack of male participants in the included studies; only two male participants were included in our review.
Relatedly, only 12% of victims included in our review were from non-Western countries. Therefore, an important aim for future research is establishing if the present results generalize to male domestic violence victims, and victims from non-western countries.
The studies included in this review were deemed to be of sufficient to excellent quality, however, all of the studies had some level of attrition bias and selection/sampling biases were common. Future studies should aim to address these biases.

| CON CLUS ION
In summary, the present study strengthened the research base on disclosure of domestic violence in healthcare settings by providing a comprehensive, updated systematic review of qualitative research barriers and facilitators of disclosure, advice of victims on disclosure, and the potential outcomes of disclosure. Results were partly consistent with a prior review, but also demonstrated novel themes and issues relevant to disclosure in healthcare settings.
The fact that barriers to disclosure persist despite the development of international guidelines and regulations for healthcare services suggests that improving HCP and healthcare service responsiveness to domestic violence remains an important goal for the future.

CO N FLI C T O F I NTE R E S T
The authors declare no conflicts of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available in Table 1 and the Supporting Information of this article.