Making the links between domestic violence and child safeguarding: an evidence-based pilot training for general practice.

We describe the development of an evidence-based training intervention on domestic violence and child safeguarding for general practice teams. We aimed - in the context of a pilot study - to improve knowledge, skills, attitudes and self-efficacy of general practice clinicians caring for families affected by domestic violence. Our evidence sources included: a systematic review of training interventions aiming to improve professional responses to children affected by domestic violence; content mapping of relevant current training in England; qualitative assessment of general practice professionals' responses to domestic violence in families; and a two-stage consensus process with a multi-professional stakeholder group. Data were collected between January and December 2013. This paper reports key research findings and their implications for practice and policy; describes how the research findings informed the training development and outlines the principal features of the training intervention. We found lack of cohesion and co-ordination in the approach to domestic violence and child safeguarding. General practice clinicians have insufficient understanding of multi-agency work, a limited competence in gauging thresholds for child protection referral to children's services and little understanding of outcomes for children. While prioritising children's safety, they are more inclined to engage directly with abusive parents than with affected children. Our research reveals uncertainty and confusion surrounding the recording of domestic violence cases in families' medical records. These findings informed the design of the RESPONDS training, which was developed in 2014 to encourage general practice clinicians to overcome barriers and engage more extensively with adults experiencing abuse, as well as responding directly to the needs of children. We conclude that general practice clinicians need more support in managing the complexity of this area of practice. We need to integrate and further evaluate responses to the needs of children exposed to domestic violence into general practice-based domestic violence training.

What is known about this topic

Although there is considerable research evidence associating domestic violence with poor outcomes for exposed children, we need mechanisms for linking them in policy and practice in healthcare settings.

There is scant understanding of how general practice responds to the needs of children exposed to domestic violence.

What this paper adds

While general practice clinicians are fully aware of their child safeguarding responsibilities, they are uncertain about best practice at the interface between child safeguarding and domestic violence.

The lack of relevant training contributes to failures to translate child safeguarding knowledge into safe and effective domestic violence‐related practice strategies.

General practice clinicians need relevant training and support in responding to domestic violence in families.

Introduction

Domestic violence damages physical and mental health (Ellsberg et al. 2008, Feder & Hester 2015) resulting in increased use of health services by survivors of abuse. The prevalence of domestic violence among women attending general practice, as with other clinical services, is higher than in the wider population (Feder et al. 2009, Britton 2012). Women experiencing domestic violence, who are often isolated from other services as a result of their partner's controlling behaviour, are more likely to be in contact with general practice than with other agencies (Hegarty 2006). Although they tend not to disclose spontaneously to their GP, they have an expectation, often unfulfilled, that doctors can be trusted with disclosure, and can offer them safe, non‐judgemental and practical support (Feder et al. 2006).

While knowledge of the impact of domestic violence on health is increasing, there is considerable scope to enhance clinicians' ability to respond appropriately to affected families (Bradbury‐Jones et al. 2011, Radford et al. 2011, García‐Moreno et al. 2015). The subject of domestic violence is virtually absent from UK medical and nursing undergraduate and postgraduate curricula and has a patchy presence in continuing professional development (Department of Health 2010, NICE 2014). Despite international (WHO 2013) and national (NICE 2014) guidelines in place on the healthcare of women experiencing abuse and the commissioning of a general practice training and support programme (Feder et al. 2011) in many areas of the UK, the majority of primary care clinicians still do not receive any formal training about domestic violence (Ramsay et al. 2012).

Domestic violence adversely affects the development, educational attainment and mental health of children (Antle et al. 2007, Chang et al. 2008, Holt et al. 2008, Stanley 2011). There is also an overlap between childhood exposure to domestic violence and other types of child maltreatment (Sharpen 2009, RCGP/NSPCC 2011, GMC 2012), but this link is scantily addressed in mandatory child safeguarding training for general practice clinicians. The RESPONDS (Researching Education to Strengthen Primary care ON Domestic violence and Safeguarding) study aimed to establish an evidence base for training on the interlinked issues of domestic violence and child safeguarding, developing and piloting a new training intervention for general practice teams. Integrating training on domestic violence and child safeguarding offered a means of improving the knowledge, skills, attitudes and self‐efficacy of general practice clinicians in managing the complexity of domestic violence when children are affected.

This paper describes how research evidence informed the development of the training intervention and identifies areas for practice and policy improvement, as well as future research directions. Detailed research methodologies and findings (Larkins et al. 2015, Szilassy et al. 2015a, Szilassy et al. 2015b, Turner et al. 2015, Drinkwater et al. 2017) as well as the piloting and outcomes of the training (Lewis et al. 2017) are reported elsewhere.

Methods

The multidisciplinary research team integrated heterogeneous evidence sources into the development of a training intervention (Figure 1). Data were collected between January and December 2013. Evidence sources included (i) a systematic review of training interventions to improve professional responses to disclosure of domestic violence when children are exposed and to identification of child maltreatment when domestic violence is present; (ii) mapping of the content of current domestic violence and child safeguarding training available in England; (iii) qualitative assessment of general practice responses to domestic violence in families with children analysing examples of positive practice and barriers to engagement; (iv) a two‐stage consensus process with a multi‐professional stakeholder group including experts on domestic violence, health and safeguarding. Integrated findings informed the design (format and content) of the training intervention which has been piloted and evaluated in England (Lewis et al. 2017).

Figure 1

RESPONDS study process.

The study was guided by two panels of professional and service user experts. It was approved by University of Bristol Ethics Committee and was conducted in accordance with clinical commissioning groups' research governance requirements.

Research evidence streams

(i) Systematic review of training interventions

We searched both peer‐reviewed and non‐peer‐reviewed international literature without any restrictions on language or study designs. We included any type of intervention or significant change in policy or practice intended to facilitate and improve professionals' response to disclosure of domestic violence in families with children and improve professionals' responses to child maltreatment in the context of domestic violence. Twenty‐one studies met the inclusion criteria: 3 randomised controlled trials and 18 pre‐post intervention surveys. There were 18 training and three system‐level interventions. We completed a narrative synthesis of these studies taking into account study design, quality, size, direction and significance of observed effects and consistency of finding. (See Turner et al. 2015) for full details of methods and PROSPERO registered protocol (http://www.crd.york.ac.uk/prospero; registration number CRD42013004672).

(ii) Training curricula mapping study

We reviewed the content of training materials on domestic violence in relation to child safeguarding in England. We mapped the range of training materials currently offered to general practice professionals in terms of their content, learning outcomes, delivery methods and target audiences. We also determined the extent to which these training materials addressed the interface of domestic violence and child safeguarding.

We contacted 250 training providers between January and April 2013 and received 32 completed questionnaires and 22 examples of training materials with some reference to domestic violence or specifically focusing on domestic violence. The diversity of materials together with variations in training delivery (level, length, target audience) limited analysis, as it was impossible to compare course contents. We therefore confined the analysis to the extent to which the sampled training materials engaged with the interface of domestic violence and child safeguarding. We assessed the materials on a 4‐point scale from ‘very good mention’ to ‘no mention at all’. We also identified a range of core and peripheral themes, approaches, learning outcomes and a range of often or rarely used teaching/learning instruments and handouts. The materials were classified independently by two researchers and results were compared and discussed. A third researcher was consulted where there were disagreements (Szilassy et al. 2015a).

(iii) Interview study

This explored direct and indirect responses to disclosure of domestic violence when children are involved and the challenges general practice professionals face in this area of practice.

A multidisciplinary academic research team conducted qualitative semi‐structured telephone interviews between May and December 2013 with 69 general practice professionals (clinical and non‐clinical staff). As the general practice response to domestic violence often emerges in the context of practice‐level work and our training intervention was aimed at general practice teams, not individual clinicians, we interviewed three key professional groups within general practice: general practitioners (GPs, N = 42) and directly employed clinical practice staff (practice nurses/PNs, N = 12) who have contact with patients registered in the practice; and practice managers (PMs, N = 15) who have a key role in implementing policy with regard to data sharing, documentation and training (Table 1). The term ‘general practice clinicians’ refers to clinical staff (GPs and practice nurses) participating in the interview study. A mix of metropolitan, urban and semi‐rural practices was recruited by email from across six areas with both high and low levels of specialist domestic violence service provision, in the north, south and midlands of England. Individual interviewees were recruited from selected practices directly (via phone/email) by the researchers and through practice administrators. General practice professionals provided verbal (audio‐recorded) informed consent for interviews. Working with a professionally and geographically heterogeneous sample contributed to a better understanding of the different perspectives of domestic violence in families with children and the barriers to engagement in interagency work.

Table 1

Interview research participants

	GPs (42)	Practice nurses (12)	Practice managers (15)
Gender
Male	17	0	4
Female	25	12	11
Age range (years)
21–34	8	2	0
35–44	11	0	2
45–54	15	8	7
55–64	5	1	4
Not known	3	1	2
Experience managing domestic violence (number of cases)
More than five	5	0
A few	13	1
One	0	2
None	18	8
None, but aware of case at surgery	6	1
Domestic violence service provision
Sparse	16	6	6
Established	26	6	9
Location
Metropolitan	11	3	5
Urban	16	5	6
Semi‐rural	15	4	4
Region
North	14	3	4
Midlands	7	4	4
South	21	5	7

Interviews explored practices in response to disclosure of domestic violence in families, recording, referrals and interagency communication. A profession‐specific vignette facilitated exploration of different professionals' views. Interviews were audio‐recorded, transcribed verbatim, loaded into qualitative data analysis software (NVivo) and analysed thematically (Bryman et al. 1994) using a coding frame incorporating concepts that emerged from the data.

(iv) Consensus process

We aimed to reach consensus on a range of controversial statements in relation to domestic violence and child safeguarding informed by studies (i)–(iii) using a two‐stage modified Delphi consensus process (Dalkey & Helmer 1963). The process took place between July and October 2013 and included a two‐stage survey focusing on contentious and ambiguous areas of practice and a consensus meeting with 28 expert practitioners and researchers representing UK general practice, safeguarding and domestic violence sectors. The findings of the survey were fed back to those participating in the meeting. They were invited to discuss the survey results and repeat the process of scoring statements consequent to the discussion (Szilassy et al. 2015a).

Intervention development

The data from the different evidence sources were integrated using a framework of five themes: making links between child safeguarding and domestic violence; engaging with victims, children and perpetrators; interagency collaboration; confidentiality and safety; and effective and acceptable training. These illuminate current organisational and attitudinal barriers as well as facilitators of good practice in respect of general practice responses to children affected by domestic violence. They also identify specific areas for practice improvement that the training intervention was designed to address.

Findings

The five key themes identified from the different evidence sources are used to structure the reporting of findings.

Making links between child safeguarding and domestic violence

Our interview study indicates substantial variation between general practice clinicians in their perceptions of the nature and strength of connections between domestic violence and child safeguarding. Although the majority of interviewed clinicians had no difficulty establishing a link in theory between domestic violence and the potential harm it represented for children, about one‐third of practitioners only made this link when prompted by the interviewer. Moreover, more than half of GPs and nearly all practice nurses said they would not necessarily make a link between child protection concerns and the possibility that domestic violence might be an issue in a family. Some of the reasons given for not exploring the possibility of domestic violence when there were known child protection concerns included domestic violence not being ‘first on your radar or list of things to ask about’ (GP31), the problem of ‘finding the time to do [it] all’ (GP28), concern that it was a ‘difficult conversation to have’ (GP26) and the assumption that children's social services would already be in contact with the family.

Our results are consistent with previous research findings (Tompsett et al. 2010, Woodman et al. 2013) showing that while GPs have no difficulty prioritising the interests of children and are familiar with the child protection procedures in an emergency, they are uncertain about the course of action when concerns are less immediate. ‘It's the ones in the middle that I struggle with’ – noted one GP (GP24). In contrast, direct violence towards a child and the young age of the children involved were both identified as risk factors that would trigger a child protection referral.

Clinicians also had concerns about maintaining a positive relationship with an adult victim when referral to a safeguarding team was likely. The need and legitimacy of breaking confidentiality to inform social services when a child was at risk of harm was broadly understood, but thresholds for referral varied. Some clinicians had strategies for managing confidentiality including practice policies, consulting the patient and routinely asking to see patients alone. However, not all those interviewed were aware of the need for strict confidentiality with regard to domestic violence.

The training curricula mapping study found that the domestic violence focus in the child safeguarding training materials entailed one or two brief mentions during a generic presentation. This was usually limited to knowledge of policies and procedures for child protection. Discussion on the needs of the parent experiencing abuse was typically missing or minimal. Nor did the training materials address the tension between maintaining confidentiality and safety for the victim while also responding appropriately to potential harm for children. All clinicians interviewed had received mandatory child safeguarding training. In contrast, only three GPs reported having received specialist domestic violence training. Failures to link child safeguarding to possible domestic violence during the interviews highlighted a key gap that needed to be addressed in our training intervention.

Engaging with victims, children and perpetrators

Most of the clinicians interviewed demonstrated a lack of confidence and experience in holding conversations about domestic violence with patients and their families. Clinicians appeared more inclined to engage directly with abusive partners than with their children. Children and young people experiencing domestic violence were rarely directly engaged. This lack of engagement with children is in striking contrast to international guidelines on child safeguarding (GMC 2012) which state doctors working with children and young people have a duty to listen and talk directly to them and to take account of children's wishes when making judgements about their best interests (p. 16).

General practice clinicians tended to assess children's needs and experience through a proxy adult, usually their mother. One GP had to correct himself to even concede that children are patients in principle: ‘We'll probably not [talk to the] children because they're not … [pause] … well they are patients’ (GP29). In contrast, perpetrators (when they were known patients in the practice) were seen as competent informants, with potential for accepting advice and support and achieving behaviour change. Some clinicians expressed concern about their lack of competence in communicating directly with children, often seeing this as a specialist role which was the remit of child health specialists or services. Talking about violence was seen as particularly difficult, even for those who had skills in discussing sensitive issues. ‘I talk to children a lot about their parents dying and things. And I find that a lot easier, funnily enough, than talking to them about violence’ (GP03). Lack of time was perceived as a barrier to working with children, as was children's lack of direct access to health services. Lack of time to engage with perpetrators was not mentioned.

Guidance on or reference to working with children and young people and talking to them about domestic violence was absent from the training materials we reviewed. Guidance on working with perpetrators and any material on confidentiality and conflicts of interests between protecting children and sustaining relationships with different family members were also missing from the training assessed.

Members of the consensus group agreed that engaging directly with children experiencing domestic violence relied on recognising them as patients and offering them opportunities to see clinicians on their own and to establish or build up existing relationships with primary care staff. The consensus meeting concluded that training should be designed to encourage appropriate direct engagement with children experiencing domestic violence and to challenge cultures of fear or avoidance.

Interagency collaboration

Insufficient understanding of the processes of multi‐agency work at the intersection of domestic violence and child safeguarding constituted a major source of frustration for the professionals interviewed. Most felt isolated from non‐primary healthcare professional groups. They described poor relationships with children's social services, characterised by lack of feedback, and limited participation in the multi‐agency child safeguarding procedures. ‘I feel that social services sometimes are this other group, this body somewhere’, said a practice nurse (PN08). Another practice nurse said ‘we [general practice and social care] don't seem to be very linked up’ (PN04). Clinicians also emphasised the absence of face‐to‐face meetings with social workers, – ‘you've no idea what they look like’ (GP09) – the lack of named people to confer with and the challenge of ‘finding the right person to pin down’ (PM07).

General practice clinicians in our sample were unfamiliar with procedures for co‐ordinating service responses to children who were below the high‐risk threshold and most did not see themselves as having a role in contributing to a ‘jigsaw’ of information about children that was shared between agencies. Some GPs relied on health visitors' access to information about families, but relationship with health visitors was described as significantly weakened in some sites due to geographical relocation. Clinicians were largely unaware of local domestic violence resources: they lacked understanding of the services available and had almost no relationship with specialist domestic violence organisations.

Clinicians recognised the importance of informal communication between professionals and regretted its absence. Communication at an individual level, reinforced by formal methods of interagency interaction, was identified as key to effective interagency work. In practice, however, effective interagency communication was limited by insufficient understanding of other professionals' and agencies' sphere of operations, as well as lack of interagency trust and self‐confidence in responding to domestic violence in families. A lack of familiarity with other agencies' policies and practices, or an absence of ‘institutional empathy’ (Banks et al. 2008), restricted clinicians' ability to gauge thresholds for child protection referral and their understanding of the consequences of referral.

The training curricula mapping study found that knowledge and attitudes to interagency partnership were only addressed in safeguarding training provided by Local Safeguarding Children Boards. However, none of the general practice professionals interviewed had attended interagency child safeguarding training. This finding is consistent with earlier research reporting negligible take‐up of interagency child safeguarding and domestic violence training by GPs in general and by male GPs in particular (Carpenter et al. 2010).

Confidentiality and safety

The review of training curricula revealed that in some localities there were mandatory policies about recording and reporting child maltreatment in cases of domestic violence. However, we identified no course content explaining how to keep appropriate records of domestic violence and there was little explicit guidance on the importance of maintaining confidentiality to protect victims of domestic violence following disclosure.

GPs and practice nurses reported diverse methods for recording both domestic violence and safeguarding concerns in patient records. The inconsistency in documentation at a national, local and practice level reflected the lack of training or guidelines on how to record. ‘To be honest we haven't had this discussion, I'm not actually sure we have a practice policy’ (GP 01) – admitted a GP.

General practice clinicians appeared particularly uncertain about how to resolve the need for both confidentiality and safety when considering documentation of abuse in the records of different family members. There were a small number of positive examples where clinicians managed these issues by discussing their strategy with the abused parent, asking for her permission to break confidentiality and then explaining how and where it would be documented. However, the majority were not confident about managing this dilemma. Clinicians were generally more confident about documenting suspected child maltreatment than domestic violence.

The consensus process highlighted the complexity of this area of practice. The effectiveness and safety of various documentation methods (with special reference to potential harms related to documenting domestic violence in the perpetrators' medical records) were questions that produced diverse expert opinions and polarised the group during the consensus survey and discussion.

These findings reveal uncertainty and confusion surrounding the best mechanisms for ensuring safety and confidentiality when documenting domestic violence. This is partly due to clinicians' lack of awareness of guidance in this area. It also reflects the lack of professional agreement on how to mitigate potential harms and apply effective safeguards when recording child maltreatment concerns in the context of domestic violence. The decision to document embodies the tension clinicians face between sharing information to promote the safety of the child, and limiting information to maintain the confidentiality and safety of the abused parent. This tension may be further complicated by the trend towards patient online access to their own health records (Woodman et al. 2015). Some clinicians interviewed were concerned about the potential for coercive partners to gain access to any records that they might document.

Effective and acceptable training

The systematic review of training interventions showed improvements in participants' self‐reported knowledge and attitudes towards domestic violence. The intervention studies also described improvements in self‐reported competence and positive change in clinical behaviour, sustained for up to a year post‐intervention. Key elements of effective interventions included an added experiential and/or post‐intervention discussion component; incorporating ‘booster’ sessions at regular intervals following training; involvement of local domestic violence agencies or other professionals with specific service expertise; and drawing from a clear and well‐articulated protocol for intervention. Multidimensionality was a key feature of the content, method and delivery of the training interventions reviewed. Programmes covered multiple topics, used teaching strategies in combination such as discussion, modelling, role‐play, rehearsal and feedback, and integrated active/passive and behavioural/instructional approaches in one session (Turner et al. 2015).

Interviewed clinicians' training preferences varied, but the majority were in agreement about the preferred format, location and training content. They clearly indicated that they would prefer face‐to‐face training delivered in their practice and all favoured short sessions (2 hours or less). While some GPs indicated a strong preference for practice‐based training for doctors only, others suggested training for the whole practice team, including administrators. Clinicians articulated a need for interactive training discussing complex real‐life cases or scenarios. They favoured training opportunities that would address the appropriate management of difficult conversations with patients, including children, about domestic violence. A third of respondents said they would like to improve their understanding of the structures and context within which social care professionals operated. ‘I think just further down the chain I'd like to know what happens rather than just my end of it’, noted a GP (GP01) clearly conveying the quest for increased ‘institutional empathy’. Informants all welcomed the idea of having input from a local social worker in the delivery of training in order to ‘know who the social workers are and what makes them tick’ (GP21). The three GPs in our sample who gave an account of having received specialist domestic violence training reported that these training events had increased their confidence about making referrals to children's social services and their willingness to discuss cases with social workers on an informal basis.

RESPONDS training

The training intervention was designed to encourage general practice clinicians to overcome barriers and engage more extensively with patients experiencing domestic violence, as well as preparing them to safeguard and support children. It aimed to encourage clinicians to adopt ‘low thresholds’ for asking questions about domestic violence and its potential impact on children and young people (NICE 2014).

Drawing on the findings reported above, the content was selected to cover the following issues: (i) linking domestic violence and child safeguarding in practice; (ii) child protection referral process and thresholds for referral; (iii) holding difficult conversations and speaking directly with children and young people; (iv) working together with other professionals and organisations; (v) record keeping, safety and confidentiality; (vi) supporting victims and the role of general practice after disclosure.

The 2‐hour training was designed for individual general practice teams delivered on practice premises. It was targeted at clinicians, but all non‐clinical practice staff were also invited to attend. Interagency working was emphasised throughout and integrated into the delivery which was undertaken jointly by two trainers, a healthcare professional and a local social work professional. In line with the findings of the systematic review, the teaching was interactive and emphasised reflection on practice. It incorporated a film which was shown in short sequences with opportunities for group discussion inserted between them. The film's narrative featured a female patient and her 10‐year‐old son. The GP modelled positive practice in asking her about domestic violence at home, and then speaking sensitively to the child on his own to elicit his experiences. The GP subsequently discussed next steps with the mother, including making a referral to children's services. The film was interspersed with short narratives from practising GPs and a social worker highlighting the challenges faced in general practice, such as lack of time in consultations, suggesting strategies for overcoming these. A follow‐up exercise included a review (or development) of individual practice teams' recording policies in the light of the training.

The training was piloted in 11 GP practices in 2014, and findings from the evaluation of this pilot are reported elsewhere (Lewis et al. 2017). The training materials, including the film are freely available at bristol.ac.uk/responds/study, although they will be developed further.

Discussion

The connection between domestic violence and child harm is recognised in UK national guidance (RCGP/NSPCC 2011), but there is scant understanding of how general practice clinicians work with domestic violence and abuse in families. Recent policy developments have highlighted general practices' key role in providing early help and intervention for children affected by domestic violence (Munro 2011, HM Government 2015), but the challenges for general practice clinicians in responding to this area of practice safely and effectively have not been addressed.

Our findings resonate with literature describing a gap between the reality and a vision that accords GPs, ‘both within government guidance and by fellow professionals, a much more pivotal role in all stages of the child protection process than they typically assume themselves’ (Lupton et al. 2001, p. 177). Consistent with recent findings (Peckover & Trotter 2015), we found a discrepancy between policy expectations and practitioner skills/capabilities in this field. The missing translation of policy into practice is reflected in the lack of training on the interface of domestic violence and child safeguarding. Clinicians are now trained to detect child abuse and they are fully aware of their child safeguarding responsibilities. However, while their roles may be more clearly defined, they lack specialised training, as well as space and time, to interact and reflect on this difficult area of work. Our findings suggest that the absence of relevant training contributes to failures to convert child safeguarding knowledge into practice strategies in the context of domestic violence.

One of the strengths of general practice is that it can respond to the needs of multiple family members, including victims and perpetrators of domestic violence and their children. It can also potentially make a key contribution to a multi‐agency whole system response at the interface of domestic violence and child protection.

Despite important recent improvements in procedures, training and guidance, our study shows that professionals still operate on different ‘planets’ (Hester 2011). The connections between ‘planets’ are limited by lack of institutional knowledge, interagency trust and self‐confidence which limit effective communication and team working. Mounting pressures on the healthcare system, increased fragmentation of child protection services (Jay 2014), cutting of domestic violence services and the lack of a cohesive and co‐ordinated approach to domestic violence, all undermine the overall effectiveness of individual responses.

The findings of this study draw attention to the low level of general practice engagement in child protection work in relation to domestic violence. While most general practice professionals recognised domestic violence as a risk factor for children's health and well‐being, the majority failed to see links between child maltreatment and the possibility of children's exposure to domestic violence. They also struggled to manage families where the risks were low to moderate (or unclear) and GPs focused on the needs of parents rather than those of children. This focus on adults (Ramsay et al. 2012) entailed a predilection for working with the abusive partner, when they were a patient in the practice, rather than with children. Our study also revealed considerable uncertainty and confusion surrounding mechanisms for recording domestic violence in families' medical records and highlighted the importance of integrated domestic violence and child safeguarding training and policies for documenting.

The poor engagement of general practice clinicians with domestic violence training and the lack of relevant training content within child safeguarding training, are currently major gaps for general practice, leading to uncertainty and resulting in missed opportunities to support victims and their children. Training gaps can lead to feelings of inadequacy and frustration (Breckenridge & Ralfs 2006, Lykke et al. 2008) and can prevent general practice clinicians recognising and responding appropriately to child harm and maltreatment. While training may be a means of improving competence and confidence in working with families experiencing domestic violence, it is important that training is appropriate and fit for purpose. This study found that completing mandatory child protection training did not necessarily lead to greater confidence in direct work with children exposed to domestic violence. Current training for general practice clinicians does not address this topic adequately; it may exacerbate fears about talking to children without highlighting the potential risks in engaging with perpetrators. It also hinders the fuller engagement of general practice professionals in this area of work by providing little guidance on effective collaborative working between general practice, children's services and the domestic violence sector.

Key strengths of our study are the integration of heterogeneous evidence sources into the development of the training intervention and the multi‐professional/multi‐agency collaborative approach emphasised during research, development and the training pilot. Another strength relates to the relatively large number and wide geographical spread of the practices and interviewees, compared with previous qualitative studies (Tompsett et al. 2010, Narula et al. 2012), enabling thematic saturation. However, participating in the interviews may have led clinicians without personal experience of domestic violence cases to contribute views based on speculation. Although adult service users contributed to analysis and development, a further limitation is the absence of children's perspectives on how and when clinicians might engage directly with them.

The main methodological limitation concerns the inclusion of a small selection of training materials in the curricula mapping study. Despite three general postings and individualised requests and reminders, we had difficulty assembling a substantial sample of training materials. The explanations for declining participation in the curricula mapping study included fears of negative evaluation in the public domain, as well as concerns about intellectual property. Despite addressing these concerns in our information sheet and further communication with the training providers, only two charitable sector organisations participated in the mapping study. The synthesis of findings across the four study components and the input from our expert groups helped to contextualise and address the deficiencies of this study component. However, the difficulties encountered during the mapping of training indicate how the creation of a commercial market in professional training can lead to reluctance to share positive practice. This finding in itself signals a line for future enquiry about the impact of commercial competition on the availability and usage of training resources. It also highlights the importance of ensuring that training packages or other outputs of commissioned research are openly available.

The complex challenges general practice professionals face in responding appropriately and safely to children exposed to domestic violence and the promising outcomes of the pilot intervention (Lewis et al. 2017) point towards the need for further research. The identification and appropriate referral of all family members exposed to domestic violence would benefit from an increased focus on the needs of children. This study suggests that general practice training on domestic violence and children could usefully be integrated with training addressing the identification of and response to both women (Feder et al. 2011) and male victims and perpetrators (Williamson et al. 2015). The feasibility, acceptability, effectiveness and cost‐effectiveness of such an integrated training programme needs to be fully evaluated.

Conclusion

We found that general practice clinicians need support in managing the complexity of domestic violence. Their skills and confidence in responding safely and effectively to adult victims and perpetrators and in talking directly with children experiencing domestic violence should be developed through appropriate training. Such training could be reinforced by supportive practice environments, improved systems of interagency collaboration, appropriate and effective documenting and improved information‐sharing systems and policies. The development and piloting of our evidence‐based training for general practice about domestic violence and child safeguarding represents a crucial first step towards strengthening the response to all family members experiencing or perpetrating domestic violence and their children.

Source of funding

This article is independent research commissioned and funded by the Department of Health Policy Research Programme (Bridging the Knowledge and Practice Gap between Domestic Violence and Child Safeguarding: Developing Policy and Training for General Practice, 115/0003). The views expressed in this article are those of the authors and not necessarily those of the Department of Health.

Conflict of interest

The authors declare that they have no competing interests.