Introduction

Domestic or intimate partner violence (DV) and sexual violence (SV) are hidden epidemics in health care settings. [1] While both men and women use violence in relationships, men use DV more frequently and severely resulting in female victims fearing for their lives [1]. SV against adults is also perpetrated by men in the vast majority. Both DV and SV are associated with serious negative health impacts, for the victim and the perpetrator. For men, using DV is associated with increased alcohol and substance abuse, rates of depression, suicidal thoughts, stress, anxiety, low self-esteem, and increased use of health services [2, 3]. For their female partners, the aftermath of injuries, fear, and stress from DV results in chronic health problems that interfere with daily functioning and quality of life. Victims experience increased rates of anxiety, depression, substance abuse, post-traumatic stress disorder, and suicidality [1]. Globally, 38% of all female homicides are DV-related, and for women of child-bearing age, DV is the leading contributor to death, disability, and illness, mainly as a result of mental ill health [4]. For their children, DV-related trauma left unrecognised is cumulative and associated with social, behavioural, emotional, and cognitive problems, persisting into adulthood [5]. Similarly, SV against women—even the more subtle forms [6]—is associated with a range of poor physical and mental health outcomes, which can persist long after the incident [6,7,8]. These include post-traumatic stress disorder, anxiety and depression [7, 8], gynaecological problems [9], sexually transmitted infections [10] and unwanted pregnancies [11].

Over the past few years, DV and SV against women have received increased attention and recognition as issues of global concern [12,13,14]. This has corresponded, perhaps not coincidentally, with an increase in survivors coming forward to tell their stories through social and traditional media. In Australia, for instance, the high-profile case of Rosie Batty, whose son Luke was murdered in 2014 by his father, was a catalyst for DV to be firmly placed on the policy agenda [15]. Batty’s subsequent advocacy profile ensured that DV remained prominent in the community’s consciousness. In many countries, high-profile cases paved the way for other survivors to come forward and tell their stories. In a similar way, global social media feminist movements such as #MeToo, #YesAllWomen and #BeenRapedNeverReported have highlighted the prevalence and impacts of sexual harassment and SV [16]. They have also provided a platform for survivors to be heard and validated. Critically, these movements advocate a shift from victim-blaming to perpetrator accountability, encouraging women to ‘call out’ the men who have assaulted them. Several high-profile cases of SV perpetrated by powerful, white men such as Harvey Weinstein have demonstrated this to great effect.

While this increased societal awareness has succeeded in making some inroads into changing the justice system, health services have lagged behind in responding to DV and SV [17]. This gap is problematic given that at least 80% of women experiencing DV seek help from health services [1], and women experiencing SV are more likely than other women to have more doctor visits in a year [18]. General practice, antenatal, mental health services and emergency departments are key places for early intervention, as health practitioners are the major professional group to whom women want to disclose [19,20,21]. International guidelines are available for health practitioners on how to identify and respond to women who have experienced DV and SV [22, 23]. However, only a minority of women and children exposed to violence are recognised in health care settings [24], and there is evidence that health practitioners often lack the essential skills and experience to respond appropriately [22]. Furthermore, addressing the victims of violence is only one part of the problem. A recent systematic review [25•] shows there is a paucity of evidence from health care settings to help men stop using DV and even less on perpetration of SV. We acknowledge that men can be victims and that DV occurs in same-sex relationships; however, the majority of evidence to date has concentrated on DV and SV against women, as they are the majority of victims.

The aim of this review is to discuss how primary care can identify and respond to DV and SV for both perpetrators and victims, placing this in the context of the #MeToo era. We begin by discussing the lack of evidence for intervening with perpetrators, particularly the lack of focus on SV. We then describe the evidence-based guidelines for female survivors, developed in 2014/15. We then briefly critique the randomised controlled trials over the last 3 years for female survivors attending health settings, including primary care, highlighting the lack of attention on SV. Finally, we make recommendations for research, practice and policy drawing on the learnings from social media movements.

The Evidence Gap Around Perpetrators

Although interventions to support female victims of DV and SV are critical, it has been increasingly acknowledged there needs to be a focus on male perpetrators [14]. In health settings, there is almost a complete lack of evidence around how to identify and respond effectively to men using violence. A recent systematic review [25•] of interventions for male perpetrators of DV in health settings found nine randomised controlled trials of low to moderate quality that were confined to specific clinical settings (alcohol treatment centres, veterans affairs). The authors concluded that there is currently insufficient evidence to support the effectiveness of any intervention, although the concurrent treatment of violence and alcohol abuse through psychological therapies showed some promise. Only one pre-post pilot study was undertaken in a primary care context [26], testing the effectiveness of a training intervention on the identification and referral of male perpetrators. Whilst the intervention did increase clinician confidence, it did not result in a significant number of referrals to specialist services. Similarly, a 2018 review [27] synthesising recommendations for how primary care should respond to perpetrators found only expert opinion pieces and guidelines to draw upon; no empirical studies were found. In the case of SV perpetration, to our knowledge, there are no studies that address identification and response in the primary care context or in any health setting, including qualitative or other non-randomised studies. The majority of SV perpetrator research focuses on prevention of SV through attitudinal change in the college context or bystander interventions.

Identification for DV and SV

Health practitioners are crucial to early intervention given their pivotal role in DV identification, safety assessment, response and referral [21]. A 2015 Cochrane review [28] suggests that screening and initial response by health practitioners increases identification with no marked increase in referrals or changes in women’s experience of violence or wellbeing. A recommendation was made that only in antenatal care should screening occur. The vast majority of women find screening or identification if there are symptoms or signs (case finding) acceptable if the questions are asked in a non-judgemental way [29,30,31]. Several studies have explored face-to-face versus more distal ways (paper or online) of asking about DV [32,33,34]. A 2015 systematic review of six randomised controlled trials [35] showed screening face-to-face does not significantly increase disclosure compared to self-administered written screening. However, computer-assisted self-administered screening was found to increase the odds of DV disclosure by 37% compared to face-to-face screening and 23% higher than self-administered written screening. A 2016 systematic review [36] found three out of ten tools; Women Abuse Screen Tool (WAST, Canada), Abuse Assessment Screen (AAS, USA) and Humiliation, Afraid, Rape and Kick (HARK, UK) had stronger psychometric values than the other seven, having been validated against an appropriate reference standard. There is much less evidence on screening for SV, other than in the context of DV. Although sexual violence victimisation scales do exist [37], there is limited evidence as to their usefulness in health settings. Most of the data on the prevalence of SV screening is conflated with data on “violence” more broadly, making it difficult to ascertain how common it is for health practitioners to enquire specifically about sexual assaults by a stranger, friend or acquaintance [38].

There are many barriers as to why practitioners may not screen or ask patients about DV or SV. A 2016 review reported low rates of routine screening of 10 to 20% (range 2 to 50%) across 35 studies [39]. Evidence shows that only half of health practitioners find screening acceptable [29, 40]. Some health practitioners do not see it as their role, are fearful of offending the patient, or feel they do not have the skills or sufficient time to provide an adequate response. Health practitioners are often impeded by the presence of the partner or feel unsupported by lack of training or referral options.

Response to Disclosure of DV or SV

In 2013, the World Health Organization (WHO) recommended all health practitioners be trained in a first-line response for DV and SV: Listen, Inquire about needs, Validate patients’ experiences, Enhance safety and offer ongoing Support (LIVES) [22]. The main skills that practitioners need to acquire often are to assess safety and risk and understand that not all patients are ready to take action [41]. Many may not wish to access formal support services as they do not identify as ‘DV or SV victims’ [42]. There is limited evidence as to what interventions assist with recovery from DV; however, they can be categorised into the following areas: first-line response and referral, safety planning and advocacy, psychological treatments, and mother-child interventions. The most promising interventions identified by this WHO work were nurse home visiting advocacy programs, mother-child psychotherapeutic interventions and specific psychological treatments (trauma-informed cognitive behaviour therapy (CBT)) [22]. There are very few studies drawn from primary care to support women with DV to recover [43]. One primary care study [44] found no effectiveness for an interpersonal therapy intervention to enhance social support and improve interpersonal functioning. In a larger study, general practitioners [42] were trained and supported to respond to women’s needs and deliver motivational interviewing or non-directive problem-solving techniques depending on women’s readiness to change. There was no difference in the quality of life between groups; however, there were some effects of the intervention with women reporting less depressive symptoms and increased discussion of safety with the doctor.

There is also limited evidence for effective responses to women experiencing SV, especially in primary care. In the immediate post-sexual violence crisis period, the use of Sexual Assault Nurse Examiners (SANEs) is becoming more common. SANEs primarily operate out of hospital emergency departments and provide first-line response and post-sexual assault crisis care, offering an alternative approach to the collection of forensic evidence that may be more woman-centred and trauma-focused [45]. To date, however, the evidence for their effectiveness has been lacking, with no randomised controlled trials being conducted in any setting. It has been suggested that this is because SANE programs tend to be responsive to the needs of communities and survivors, making ‘success’ difficult to measure [45]. One study found a significant increase in prosecution of SV cases post a SANE intervention compared to before implementation [46]; however, rates of reporting were incredibly low.

Beyond a crisis response, a 2015 Cochrane systematic review [47•] of controlled studies found some evidence for psychological therapies such as assertion training, clinician-assisted emotional disclosure, CBT, eye movement desensitisation and reprocessing, prolonged exposure therapy, stress inoculation therapy and supportive psychotherapy in reducing some of the mental health impacts of SV. However, with the exception of CBT and prolonged exposure therapy, the long-term benefits (beyond 12 months) of these therapies were not established. Furthermore, interventions were delivered in a variety of clinic and university settings, and it is unclear to what extent they could be successfully delivered in primary care. A systematic review in 2013 [48] found similar results, concluding that there was a paucity of rigorously evaluated psychotherapeutic interventions for SV. Trauma-informed approaches to treatment for SV survivors show some promise [49, 50], but are yet to be evaluated.

Reproductive coercion, another hidden issue, has received little attention in research and policy. Emerging only in 2010 as a topic of discussion [51], it refers to interference with a woman’s reproductive choices (typically by a male partner) and includes forced abortion and contraceptive sabotage [52]. Reproductive coercion sits within the intersection of DV and SV [53]; primary care is thus well-placed to identify and respond to it, yet almost no evidence exists to support practitioners in doing this effectively [54].

Recent Evidence

Our search over the last 3 years revealed only one primary care randomised controlled trial—a screening and brief intervention via a resource card for reproductive coercion and DV in the family planning context (ARCHES). The intervention was not superior to usual care in reducing reproductive coercion or DV [55]. Extending our search to DV or SV trials from other health settings showed mixed-quality studies that focussed on advocacy and empowerment, integrated interventions (with alcohol or drug misuse) and couples counselling for DV [56,57,58,59,60,61,62,63] and cognitive behavioural or processing therapy for SV.

Advocacy and empowerment interventions [64] have been shown to increase referrals to services and are the mainstay of response to DV. A 2015 Cochrane review concluded that there is some support for advocacy effectiveness. Intensive advocacy may improve quality of life and reduce physical abuse up to 2 years later and brief advocacy may have short-term mental health benefits and reduce abuse, particularly in pregnant women and for less severe abuse. However, in a large 2017 well-designed trial [65•], 950 women experiencing DV attending 42 public health clinics in Mexico were randomised to a nurse-delivered session (DV screening, supportive referrals, safety risk assessments) and a booster counselling session after 3 months, or to screening and a referral card. There were no significant effects observed at 6-month follow-up, although immediately after the booster session, there were significant improvements in mental quality of life and safety planning behaviours. Similarly, in a recent randomised controlled trial [66] of telephone-delivered nurse support (referrals and social support) versus enhanced usual care for 300 women who reported DV within the past year attending a paediatric emergency department showed no difference on outcome variables, including DV, depressive and post-traumatic stress symptoms. Overall, the evidence for advocacy, although the strongest, still needs to be qualified by the small number of trials of mixed quality.

In contrast to the above, a randomised controlled trial of empowerment of women (n = 239 women) in the context of perinatal home visiting in the USA showed some positive results [67]. The Domestic Violence Enhanced Home Visitation Program (DOVE) intervention group received a structured abuse assessment and six home visitor-delivered empowerment sessions integrated into home visits. Women in the DOVE treatment group reported a larger mean decrease in DV scores from baseline compared to women in the usual care group over a 24-month period. The average 5-point reduction in scores was statistically significant and a clinically meaningful difference.

Several recent trials have attempted to integrate interventions for DV and alcohol or drug misuse. A 2015 randomised clinical trial [68] at two US emergency departments assessed the effectiveness of a motivational intervention (30 min) by masters-level therapists for 600 female emergency department patients who exceeded safe drinking limits and were experiencing DV. At 12 weeks, there were no significant differences between groups for experiencing DV or heavy drinking [69]. In a small low-quality trial [70], a 10-session CBT group was piloted in Spain among 14 women receiving outpatient treatment for a drug use disorder who had DV in the previous month. The intervention did not significantly reduce the likelihood of any DV, depressive symptoms, quality of life or health status. These results do not support widespread implementation of brief interventions for addressing co-morbid DV or heavy drinking or drug misuse. Alternative settings or more intensive interventions may be needed or in the latter study larger sample sizes to detect any differences.

Although there are many recommendations for health practitioners not to deliver couples interventions [22], several researchers are testing them in specific settings with mixed results. In a low-quality trial [71] of 69 male service members or veterans and their female partners attending two hospitals, the Strength At Home Couples Program (SAH-C) was compared to “supportive prevention” couples therapy. Completion rates for both treatment arms were low, and consequently, no statistically significant between-group differences on measures of domestic violence could be considered. In a secondary analysis of a randomised clinical trial [72] comparing behavioural couples therapy plus individually based treatment versus individually based treatment, there is a suggestion that individual is better than couples.

In a previous WHO systematic review, mother-child interventions have had the strongest evidence for assisting women [22]. More recently, a community-based intervention [73], the Moms’ Empowerment Program, was tested with 181 mothers exposed to DV during the past year. Participants were allocated to three conditions: mother-plus-child intervention, child-only intervention and a wait list comparison. Women in the mother-child condition showed the greatest improvement over time in positive parenting and depression.

Very few randomised controlled trials were found addressing SV identification or response in health settings over the last 3 years, and the results of these were mostly inconclusive. One 2017 study compared cognitive processing therapy (CPT) to usual care in 47 women experiencing sexual assault-related acute stress disorder [74]. Small between-group effect sizes were reported in favour of CPT for both PTSD and depression, although the authors were unable to conclude that CPT was meaningfully superior due to the small sample size and large confidence intervals. Another study [75] compared CBT plus imagery rehearsal therapy (IRT) to CBT alone for 42 sexual assault survivors experiencing PTSD. They found no statistically significant differences between the two groups on mental health outcomes or on night time distress.

Health Systems Change Is Needed

The WHO recommends a broad systems-based approach to enable sustained change in health practitioner behaviour [21]. Improving patient outcomes in the context of DV and SV requires patient-centred care, but also a ‘whole of system’ health service response. At a health provider level, changes could include promoting a culture of gender equity; having trauma-informed principles (respect, privacy, confidentiality, safety); allowing sufficient consultation time and promoting awareness of DV and SV protocols and referrals. At a systems level, change might involve provision of workforce support and mentoring, appointment of champions, allocation of finances to family violence services, and information systems for evaluation. A survey of health clinics across Europe found several factors encouraged best responses including the following: (1) committed leadership; (2) regular training with mandatory attendance that included front-desk workers to health care providers; (3) use of on-site trainers and (4) a clear referral pathway [76].

Conclusion

There is extremely limited evidence over the last 3 years to guide health practitioners in the areas of DV and SV. Consensus guidelines from the WHO remain that screening in antenatal care is warranted and that all practitioners should be trained to deliver a first-line response which includes a safety assessment. Advocacy and empowerment for DV appear to be most beneficial when delivered to women in the perinatal period and through home visiting. Some patients are unable to access health care or are reluctant to disclose face-to-face as they fear judgemental attitudes, and so, researchers testing online safety decision aids and healthy relationship tools are showing acceptability, feasibility and some efficacy [57, 58, 60, 62, 63, 77]. For SV, there is older evidence to support the use of CBT and prolonged exposure therapy as an effective response, although not specifically in the primary care setting. No randomised controlled trial evidence could be found for effective ways for health practitioners to identify or ask about SV in health settings.

There are a number of critical gaps in the literature that our review has identified. First, the almost complete lack of evidence to support effective ways for practitioners to identify and respond to male perpetrators of DV and SV. Second, the lack of robust evidence to guide practitioners in identifying and responding specifically to SV, in primary care or more broadly across health settings. Although there are overlaps between DV and SV, it is not enough to address SV simply as part of a DV response; this does not acknowledge the particular nuances of SV perpetrated by an intimate partner, nor does it account for SV perpetrated by strangers [78]. How interventions might work for different groups of SV survivors is unclear [48]. Third, there is a need for a tailored response to women, men and their children as DV experience is heterogeneous, with patients at different stages of readiness to take action [17, 79].

Considering the review findings in the context of #MeToo and other social media movements highlights some interesting contrasts. These movements have highlighted SV and DV and the need to listen to survivors and make men accountable for their actions. It has been suggested in the news media that this has resulted not only in increased awareness about violence against women but also galvanised support for survivors and led to repercussions for perpetrators. A recent New York Times analysis, for instance, reported that over 200 powerful men had lost their jobs as a result of perpetrating sexual harassment in the workplace, with many of them being replaced by women [80].

If we look specifically at the health sector, however, it appears clear that #MeToo has not impacted on how SV and DV are addressed. We are yet to see an emphasis on perpetrator accountability in the health sector, nor even attempts to promote identification and referral. Furthermore, despite the primary focus of the #MeToo movement being SV, there is a dearth of literature around how to address this issue in health settings. The majority of health settings research focuses on DV, whose corresponding social media movements #WhyILeft and #WhyIStayed have not taken off in the same way as #MeToo [81]. It appears, then, that social media movements may not be the most effective way to encourage health systems change; new and innovative ways of translating social activism into concrete action for the health sector is critical for the future.