1 Introduction

Identifying and operationalizing ‘good practice’ in child and adolescent-to-parent violence and abuse (CAPVA) remains challenging, in part because CAPVA is defined and conceptualised inconsistently across research, policy, and practice. In this paper, CAPVA is used as an umbrella term referring to a pattern of repeated harmful behaviors (including, but not limited to, physical, psychological, emotional, and coercive/controlling behaviors) enacted by a child or adolescent toward a parent or primary/secondary caregiver within a family context. CAPVA is a complex phenomenon in which intentional or unintentional harm may be repeatedly inflicted on parents over a sustained period. This complexity is underpinned by multifaceted interactions between the social, individual, and biological factors (Simmons et al. 2018), including distress associated with unmet needs and/or maladaptive strategies for meeting needs (Rutter 2023, 2024). Regardless of cause, families frequently do not receive support until harm escalates to crisis, with a risk of child removal or child criminalisation (Holt and Lewis 2021; Rutter et al. 2025).

Ideally, wraparound and multi-agency working would address the complex determinants associated with CAPVA and reduce the likelihood of the criminalisation of young people; however, such approaches are infrequently implemented in practice (Davies and Walker 2025, 12). Conceptualising children as victims while simultaneously recognising their use of harmful behaviour has been advocated, on the basis developmentally inappropriate behaviour may be harmful to the child as well as to others (Rutter et al. 2025; Condry and Miles 2021; Hackett et al. 2006). Nevertheless, framing children as victims of their own harmful behaviour has also been argued to increase the risk of attributing responsibility to parents for an alleged inability to control their children (Coogan 2011).

Existing scholarship demonstrates substantial heterogeneity in how CAPVA is framed (as domestic abuse, family violence, safeguarding, neurodevelopmental/mental health need, or trauma-related behaviour) and, consequently, in recommended service responses. Recent evidence syntheses similarly highlight definitional ambiguity and variability in intervention models and practice guidance, alongside limited agreement about effective pathways to support (Baker and Bonnick 2021; Rogers and Ashworth 2024; Toole-Anstey et al. 2023). This ambiguity provides the rationale for a structured consensus-building approach. The Delphi technique is well suited to contexts characterised by emerging evidence and contested concepts, because it provides an iterative, anonymised mechanism for eliciting, refining, and quantifying expert judgement (Mikton et al. 2017). While Delphi methods are often used to build consensus across fragmented fields, the purpose of this study is not simply to aggregate expert opinion. Rather, by deliberately bringing together practitioners, researchers and lived experience specialists who work with fundamentally different presentations of CAPVA; this paper addresses a core analytical problem in the CAPVA field: the persistent misspecification of heterogeneous family experiences as a single, unified phenomenon. Thus, we advance the argument that the continued search for uniform definitions, thresholds, or pathways has itself become a limiting assumption, one which this paper directly addresses by establishing principle led consensus designed to operate under conditions of heterogeneity rather than to eliminate it.

1.1 Framings

CAPVA researchers most commonly apply domestic abuse frameworks when interpreting data relating to CAPVA (Simmons et al. 2018; Rutter 2023). However, this appears to be due to victimisation similarities between CAPVA and domestic abuse around ‘dilemmas of disclosure’ including parental ‘minimization’ of the harm, and self ‘blame’ (Coogan 2011, 355). Contrary, other researchers have highlighted any conflation of CAPVA with domestic abuse blurs conceptual and practice boundaries (Holt and Lewis 2021). Furthermore, there are substantial practice-based problems in applying domestic abuse frameworks to the harm instigated by children, which include criminalising children, reluctance to report, responsibilising parents and children, and stigmatising families (Miles and Condry 2015). Indeed, any practice which considers children ‘perpetrators’ of abuse, rather than children in need of support has been associated with direct harm to children, preventing them from establishing healthier behaviors (Miles and Condry 2015; Venables et al. 2025). There is also high association with developmental differences in children who instigate CAPVA, leading it to be widely associated with child development challenges rather than a criminal issue (Holt 2016; Miles and Condry 2015; Rutter 2022; Rutter et al. 2025).

It is important to avoid replacing one dominant framing with another. CAPVA is best understood as a heterogeneous set of harmful behaviors and relational dynamics, arising in diverse family contexts and through multiple, sometimes overlapping, pathways (trauma exposure, neurodevelopmental difference and sensory dysregulation, cumulative adversity, parental ill-health, family conflict, and broader structural stressors (Baker and Bonnick 2021)). This creates additional complexity regarding good practice guidance, as it is highly contextual.

1.2 Relevant Legislation

The Domestic Abuse Act 2021 was a milestone moment in UK legislation, solidifying for the first time the role of a domestic abuse commissioner, recognising children as victims in their own right, whilst outlining anyone of the age of 16 could be instigating domestic abuse (Domestic Abuse Act 2021). However, it is acknowledged children under the age of 18 should be provided with a safeguarding response (Working Together to Safeguard Children 2015). That the harm caused by those aged 16–18 could be included in this act led to an English expansion of funding available to support those living with CAPVA, in both 2021 and 2022 Durham County Council received this funding for families with children aged eight-to-18 who were instigating CAPVA, to provide training for professionals and interventions for families. Over the first 2 years they reported the following:

In the first year, training was delivered to 230 professionals while 800 relevant professionals were engaged in awareness raising sessions. Additionally, a victim safety worker supported 11 families. In the second year, the support worker supported 48 families. Funding was also invested in further training for team leaders and the incorporation of the Respect Young People Programme into the school curriculum.

Accordingly, this study examines the perspectives of localised practitioners who have participated in CAPVA-related training in the context of recent legislative developments in 2025 and considers these alongside the views of experts internationally who identified as experts in a 2021 study. Whilst these are discrete studies and therefore this is not a paper of change over time, examining the perspectives of two distinct panels enables identification of areas of convergence and divergence in how CAPVA is conceptualised and how ‘good practice’ is operationalised across contexts. This comparative lens is important because it illuminates which principles appear robust across different settings and which are contingent on local service configurations, resources, and policy environments, thereby strengthening the practical utility of the consensus findings for service development and guidance.

2 Method

This paper reports findings from two independent Delphi studies (Delphi 1: 2021; Delphi 2: 2025). The overarching objective was to identify areas of convergence and divergence in how CAPVA is understood and framed, and what constitutes appropriate assessment and support, including referral processes and service delivery.

2.1 Ethical Considerations

Both procedures were approved by Durham University, with Delphi 1 receiving approval in 2020, and Delphi 2 receiving approval in 2025. In both Delphi studies, a link was provided to JISC surveys, which hosted the Delphi questionnaires, whereby experts could see an information sheet including recruitment criteria for ‘experts’, as well as information regarding right to withdraw, and the data management plan.

2.2 Participants

2.2.1 Delphi 1 (2021)

Delphi 1 used purposive sampling, consistent with Delphi guidance (Hackett et al. 2006). Invitations were distributed to specialist CAPVA services and subsequently broadened via social media, including a post on the ‘Holes in the Wall’ CAPVA information website. Eligibility criteria were designed to ensure substantive topic expertise, while also incorporating lived experience to reflect the participatory orientation of the wider program of work.

In Delphi 1, participants were eligible if they met at least one of the following criteria:

• Specialist practitioners who were working with young people initiating CAPVA, and these behaviors began before the age of 12.
  or
• Specialist CAPVA practitioners who have been working with a young person initiating CAPVA and there is a younger sibling in the home (under the age of 12) who had also been identified as presenting with similar behaviors.
  or
• Parents of children/adults aged over 16 where their children exhibiting relevant behaviors prior to age 12.

Some experts fit into both practitioner and parent categories; therefore all experts were asked to identify if they were an expert-practitioner, expert-by-experience, or both. Experts by experience have successfully been participants in health research using the Delphi approach (Law and Morrison 2014), and they were included in Delphi 1 to reinforce the participatory principals underpinning this project. Years of practice were not required, and participation was open from any country.

Table 1 summarises the number of respondents per round. In classical Delphi designs, a stable panel is ideally retained across rounds to examine how expert judgement changes in response to controlled feedback. Delphi 1 used a modified approach in which the Round 1 panel was expanded for Round 2 to maximise the breadth of specialist and lived-experience input at the point where statements were first rated. This recruitment approach is a common methodological modification (McMillan et al. 2016). This was achieved by re-contacting all Round one invitees and circulating the Round 2 survey more widely within the same sampling frame. As a result, the number of respondents increased from Round 1 to Round 2. To preserve the longitudinal element of the Delphi design. Round 3 was then restricted to Round 2 respondents, enabling within-panel comparison of re-scoring following controlled feedback.

2.2.2 Delphi 2 (2025)

Delphi 2 was conducted as a separate, locally situated Delphi study in collaboration with an Durham County Council local authority in the north-east of England. Participants were recruited via the partner organization and comprised practitioners involved in delivering CAPVA-related intervention and/or providing CAPVA-related training to other professionals. The Delphi 2 panel therefore reflected practice-based expertise within a single service system, rather than an international cross-sector expert sample, participants and their service are available (Table 2).

2.2.2.1 Advisory Group (Delphi 2)

In addition to the Delphi 2 practitioner panel, an advisory group was convened to support co-design and interpretation. The group comprised of eight service leads and representatives, and three parents with lived experience of CAPVA who were navigating local pathways to support. The advisory group contributed to the development of Round 1 prompts (to ensure relevance to local referral processes and service-access challenges) and met again after completion of the Delphi rounds to discuss the findings, with a specific focus on feasibility and implementation within the local system. Advisory group contributions were collated utilising a workshop approach, and used to contextualise and elaborate Delphi 2 findings; however, advisory group views were not included in the quantitative consensus calculations.

2.3 Instruments

Each Delphi study comprised three sequential questionnaires administered online. Delphi 1 was initially constructed based on criteria set out in Rutter (2021), whereas Delphi 2 was initially constructed by an advisory group of service leads and parents navigating CAPVA after they were presented with the findings of a participatory action research project on service access (Rutter et al. 2025). The advisory group constructed questions for Round one of Delphi 2 and then met again once the Delphi procedure was completed to discuss the findings.

2.4 Procedure

Data were collected via JISC Online Surveys. The studies were conducted independently (Delphi 1: international cross-sector expert panel; Delphi 2: local practice panel) but followed the same three-round structure, with each round open for 6 weeks (round 1 and round 2 questionnaires for both studies are available in Supporting Information S1). Round 2 converted Round one qualitative responses into a structured set of statements for rating (Delphi 1: 46 statements; Delphi 2: 36 statements). Statements were rated on a 10-point Likert agreement scale (1 = strongly disagree to 10 = strongly agree) and participants could provide a free-text rationale for each rating. A ‘don't know/not applicable’ option was included in Delphi 1, but not Delphi 2; where included, these responses were excluded from the denominator when calculating percentage agreement/disagreement. Round 3 re-presented only those items that did not meet the pre-specified consensus criteria in Round 2 (following McMillan et al. 2016). Participants received individual questionnaires with their individual responses in comparison to other participants, and qualitative responses contradicting their own response to see if they wished to change their score. The scoring criteria used to operationalise agreement/disagreement and consensus are detailed in Table 3.

2.5 Data Analysis

Quantitative analysis was conducted using SPSS. For each statement rated on the 10-point agreement scale, responses were summarised using percentage agreement/disagreement, the median (M), and the interquartile range (IQR) as an index of dispersion. Following common practice in Delphi reporting, percentage thresholds were used to indicate higher and lower levels of convergence; however, because there is no single agreed standard for Delphi consensus criteria, thresholds were decided upon after careful consideration of existing Delphi studies (Der Gracht 2012). All criteria needed to be met; otherwise the consensus level was downgraded to the next band. The full Likert scoring was used in Round 2 for both studies, but collapsed from 10 to five bands for round 3 to enforce further consensus, a process which was successfully applied in other work (Hackett et al. 2006).

3 Findings

Both Delphi studies elicited qualitative responses, facilitating anonymised dialogue among participants and providing additional contextual detail regarding respondents' perspectives. In Delphi 2, findings were also discussed with the advisory group to examine levels of agreement with the emerging consensus and to identify feasible actions to improve access to interventions. Thus, qualitative and quantitative data is presented in this findings section. A central interpretive consideration is heterogeneity. Although both Delphi studies identified convergence around broad practice principles (e.g., non-stigmatising, child-centred responses; avoidance of perpetrator framings; timely access to support), CAPVA presentations are likely to vary markedly by developmental stage, severity and chronicity, co-occurring needs, and family ecology. Accordingly, the present findings should not be read as endorsing any single causal account (e.g., ‘CAPVA as trauma’, ‘CAPVA as neurodevelopment’, or ‘CAPVA as domestic abuse’), nor as implying one intervention package will be appropriate across cases. Instead, the consensus statements can be understood as an applied framework for developing flexible pathways which can be adapted to local systems while maintaining core principles: safety, relational engagement, and proportionate support for both parents and children.

3.1 Delphi¹

High consensus agreement in relation CAPVA knowledge including: ‘CAPVA is more widespread than we realise’ (97.8%, 0, 1); ‘We should always see the child before the behaviour’ (91.1%, 0, 1. Furthermore, the Delphi 1 statement ‘CAPVA is a public health issue’ (89.7%, 1, 1) connected to Delphi 2 as whilst it was not included in the Delphi 2 questionnaire, but the advisory group proposed CAPVA should be framed as a ‘public health crisis’ and ‘be the center of public health campaigns’, demonstrating how both at the international and local level, public health has a role to play in reducing and/or preventing CAPVA.

Also achieving high consensus agreement at Delphi 1 were the statements: ‘Children do not want to be violent and are also victims’ (81.8%, 1, 1); ‘Children under 11 initiating violence should be understood through a trauma informed lens’ (79%, 1, 1); ‘Perpetrator models should be avoided with children’ (78.8%, 2, 2). This consensus agreements highlight the importance of providing de-stigmatising, child-centric approaches to intervention. Furthermore, Experts could not reach agreement on the statement ‘CAPVA is a child trying to gain control of their parents’ (55.3%, 3, 3), with qualitative responses highlighting divergence regarding nuance and heterogeneity regarding CAPVA reasons such as ‘sometimes it is about a child trying to gain control because they feel out of control’ (qualitative response).

There were also a number of statements relating to early identification which achieved high consensus, including: ‘Using violence toward animals is indicative of using violence in the home’ (74.4%, 1, 2). High consensus disagreement in this section disagreed ‘damaging property is not a problem and can help children direct their anger in a healthy way’ (76.9%, 1, 5). However, there was no agreement on the statement ‘Having an intense interest in horror, murder, killings, can be indicative of a problem’ (60%, 3, 2.5).High consensus agreement regarding pathways to support included: ‘All professionals who encounter children and parents should be trained to identify CAPVA’ (93.2%, 0, 1); ‘Existing safeguarding pathways which are meant to keep children and adults safe from abuse are not adequate in cases of CAPVA’ (89.1%, 1, 1); ‘If calling the police is the only option, many families will keep CAPVA hidden’ (76.9%, 2, 1); ‘Distinct referral pathways should be available for CAPVA, where CAPVA specialists collate information from agencies and complete all direct work’ (70%, 1, 1). Thus, supporting multi-agency awareness and collaboration, rather than single-agency responsibility.

There was high-consensus agreement ‘a clear understanding of family dynamics is crucial before beginning work’ (88.9%, 0, 1); ‘assessments for sensory and neurological differences can be helpful for families’ (82%, 1, 1); ‘assessments should be completed with the whole family rather than fragmented’ (81.3%, 1, 2); and ‘parent accounts should be prioritised’ (70%, 1, 1). Within this section, there was high-consensus disagreement with the proposition ‘most families seeking help do not require support because children will grow out of the behaviors without intervention’ (72.5%, 1, 5). ‘So-called common sense/traditional parenting does not work if there is violence’ (71%, 1, 1) however ‘punishment and rewards do not work and can escalate violence’ (76.9%, 1, 1). Thus, parents may become more authoritarian in their approach when traditional strategies do not work, resulting in an increase in harm within the home overall.

When working with families, there was high-consensus agreement ‘CAPVA is not caused by a failure in parenting’ (84.6%, 0, 1); ‘it is crucial to work with families for prolonged periods to be effective’ (79.5%, 1, 1); ‘peer support is essential, so group sessions are the preferred way to support families’ (76.4%, 1, 1). ‘Safety planning and risk assessments should be completed with every family’ (74.4%, 1, 1). There was high-consensus disagreement with the statement ‘work should not include a social worker under any circumstances’ (70.3%, 2, 5).

3.2 Delphi²

When working directly with families, there was high-consensus agreement ‘asking parents if they are coping should happen at first contact and continue throughout involvement’ (92%, 0, 5; Delphi 2). The Delphi 2 advisory group described this as ‘good practice’ and emphasised the importance of regular ‘check-ins’. There was also consensus ‘professionals who work with families should be informing families they can self-refer for CAPVA’ (Delphi 2). However, this held tension in the advisory group because most referrals into services were already via self-referral, and the group cautioned this could ‘put all the responsibility back on the parent’ (advisory group). Additional high-consensus statements in this section included: ‘observing children's behaviour across multiple environments (home, school, community) assists professionals in identifying behavioural challenges, validate parental concerns, and plan tailored support’ (76%, 0.5, 5); ‘clearer referral criteria and guidance is needed to support families and professionals to navigate First Contact processes’ (88%, 1, 5); and ‘providing a private one-to-one environment to facilitate discussion of sensitive issues’ (93%, 0, 5).

A Delphi 2 statement indicated ‘help should be offered as soon as families ask for it’ (Delphi 2; Rutter et al. 2025). Across both Delphi studies, participants emphasised the importance of believing family reports and centring parental accounts. Recognising parents as experts appears to be a central component of relational practice.

‘CAPVA expertise must go beyond traditional domestic abuse models, especially when supporting children with long-term support needs’ (100%, 0, 5); ‘Families should receive tailored support and regular check-ins from a named professional following a diagnosis from CAMHS [child and adolescent mental health services], not just a written outcome or discharge’ (87%, 0, 5); ‘All professionals who work with families should be aware of the CAPVA risk assessment they can complete with families’ (84%, 0, 5).

Whilst Delphi 1 focused on those under the age of 12, Delphi 2 provided more opportunity to explore the nuance of different age groups' ‘CAPVA can be harder to identify in early years due to developmental factors and limited verbal expression in young children. Therefore, a national definition and guidance on CAPVA for children under 8 would support clearer referral pathways in early years’ (70%, 1, 5).When younger children are referred to CAPVA pathways, they may ‘require simplified and more therapeutic tools, such as play or art-based approaches which allow for emotional regulation to be the main focus of support for younger children showing CAPVA behaviors’ (advisory group). This aligns with the broader consensus ‘CAPVA interventions should be tailored to a child's developmental stage rather than their age’ (100%, 0, 5; Delphi 2). The advisory group also noted limited information-sharing between services can make it difficult to establish a child's developmental profile and needs in practice; consequently, bespoke planning and program adaptation are often required.

Barriers to accessing support was recognised as a significant concern: ‘Families experiencing long delays in appointment allocation should receive interim support, such as check in calls, whilst they are on waiting list’ (77%, 1, 5); ‘Parents with lived experience of CAPVA are experts in understanding and navigating their family's situation’ (100%, 0, 5); ‘A multi-agency approach involving professionals from various disciplines and families is essential to provide expert CAPVA support’ (84%, 1, 5).

Several Delphi 2 statements emphasised non-stigmatising communication and consistent information-sharing can reduce shame and enable earlier help-seeking (e.g., the need for honest, consistent communication and proactive guidance before formal referral points). The advisory group noted, locally, many children discussed in practice were on a neurodevelopmental assessment pathway (‘neuropathway’). This was interpreted as a contextual feature of local referral patterns rather than an explanatory account of CAPVA, but it highlighted a service-access risk: diagnostic processes can delay support and contribute to diagnostic overshadowing if harmful behaviour and safety planning are not addressed alongside assessment.

‘Professionals should receive specialist CAPVA training that includes skills for empathetic and non-stigmatising communication with parents, before completing the CAPVA screening tool’ (100%, 0, 5). However, the Delphi 2 advisory group characterised this as ‘blue sky thinking’ and emphasised school staff with a role in supporting children with special educational needs should be prioritized for such training, noting universal training for all professionals would be an ‘ideal world’ aspiration.

‘Printed materials such as posters and leaflets placed in schools, libraries, GP surgeries, and Family Hubs can effectively inform families about CAPVA’ (100%, 0, 5); ‘Social media, websites, GP surgery screens and QR codes are essential tools for raising awareness of CAPVA in the wider community’ (92%, 0, 5). Nevertheless, the Delphi 2 advisory group recognised several campaigns had occurred in the local area, but the targeting of these was unclear ‘you only seem to know where to find things if you already know about them’. Furthermore, the lack of hard copies of documents to share with families, and an unclear responsibility for why we should be maintaining these messages meant campaigns were often short-lived: ‘we need public health to take a national approach to this so people know what it is and where to go for help’.

‘Professionally designed information packs that are parent- and child-friendly should be provided to families at the point of referral to explain CAPVA and available support’ (100%, 0, 5). ‘Public awareness campaigns and peer stories can help reduce stigma and encourage families to speak openly about CAPVA’ (84%, 0, 5). ‘Improving CAPVA awareness requires clear messaging across public spaces, social media, and schools’ (100%, 0, 5). Another proposed method of raising awareness included ‘the local authority should be creating and leading groups/support circles for people experiencing CAPVA who can then become ambassadors in their own right’ (advisory group). This was developed further in the advisory discussion, where the value of sharing positive outcomes to build trust between service providers and families was emphasised. The advisory group also raised concerns about over-burdening peer ‘ambassadors’ and argued peer mentoring should complement rather than substitute the existing workforce.

4 Discussion

A central implication of both Delphi exercises is CAPVA is not a single, uniform phenomenon and should not be treated as such in either policy or practice. Existing evidence indicates substantial heterogeneity in onset, severity, duration, and the relational and contextual circumstances in which CAPVA occurs (Simmons et al. 2018; Rogers and Ashworth 2024). For some families, CAPVA appears closely connected to neurodevelopmental difference and sensory dysregulation; for others it is more closely associated with trauma exposure, cumulative adversity, family stressors, mental health needs, or conflict escalates during developmental transitions. These pathways are not mutually exclusive and may co-occur, with meaning and risk shaped by broader ecological factors. Accordingly, these present findings should be interpreted as consensus about broad practice principles, rather than as support for any single explanatory model or a one-size-fits-all intervention package.

4.1 Implications for Investigative and Statutory Responses

Although CAPVA is frequently located within family support and safeguarding systems, the Delphi findings have direct implications for investigative and statutory responses because police contact is often experienced as a default or crisis pathway when other routes to support are unclear or unavailable. Across both Delphi studies, participants emphasised families may minimise, delay, or avoid disclosure when they anticipate criminalisation of the child, parent blame, or escalation through punitive responses. This highlights the importance of designing first-contact and information-gathering processes reduce barriers to disclosure, differentiate help-seeking from offender management, and attend to the relational and developmental context in which harmful behaviour occurs.

4.2 Delphi²

4.2.1 A Public Health Crisis

CAPVA being recognised as a public health issue aligns with international scholarship that conceptualises violence as a public health problem requiring a prevention-focused response (Wolf et al. 2014). Importantly, a public health framing is compatible with heterogeneity: it supports a focus on upstream prevention, early identification, and graduated intervention, rather than presuming a single explanatory model. In this sense, the Delphi findings can be interpreted as advocating a whole-system response that is attentive to developmental needs (including neurodevelopmental and sensory profiles where relevant) while also recognising the role of trauma, adversity, family conflict, and wider contextual stressors (Holt 2023, 2024; Rutter 2024; Rogers and Ashworth 2024). In addition, the recognition CAPVA may present as a developmental concern for some families suggests specialist support can be important in preventing escalation, agreeing with work across the UK and globally that specialist intervention can support families to navigate and reduce CAPVA (Rutter et al. 2025; Curtis et al. 2022; Erostarbe et al. 2018; Toole-Anstey et al. 2023).

Despite the importance of early intervention, there are clear problems relating to both family and practitioner awareness of CAPVA support services. It appears from the respective Delphi studies that awareness raising activities are increasingly recognised as a driver in helping families access the correct and timely support. However, the lack of universally agreed terminology and definitions of CAPVA mean families are frequently, unaware of what to name the phenomenon they are experiencing, and so they have no idea where to go for help (Condry and Miles 2014).

Discussing the findings with our advisory group, it was highlighted multiple statements outlined potential options for awareness raising in Delphi 2, including printed materials (posters, leaflets, and professionally designed information packs) and digital resources (social media, websites, QR codes). However, awareness is not enough, and positive stories are the most effective in promoting positive engagement with services (Lewis et al. 2004). This is necessary as campaigns which wholly focus on the negative impacts, rather than improving knowledge about why such harm occurs, were thought to increase stigma and prevent families from seeking intervention. This correlates with existing literature, in which parents of autistic children experienced stigma due to a lack of understanding from others as to why their children may behave atypically (Broady et al. 2017).

4.3 Domestic Abuse Frameworks Are Not Appropriate

As highlighted in the introduction of this paper, whether CAPVA is domestic abuse, or should be understood through the application of domestic abuse frameworks, is highly contested (Condry and Miles 2021; Coogan 2011). The consensus from both Delphi studies indicated good practice should avoid routine application of domestic abuse frameworks or perpetrator models when working with children and families experiencing CAPVA. Participants' rationales emphasised the risks associated with criminalising children, increasing reluctance to disclose, responsibilising parents, and stigmatising families, alongside the importance of recognising children's developmental status and support needs (Holt and Lewis 2021; Miles and Condry 2015). When explaining why alternative approaches are required, respondents drew on multiple, sometimes overlapping, understandings of how CAPVA may arise, including adversity and trauma exposure, neurodevelopmental difference and sensory dysregulation, mental health needs, family stressors, and conflict that escalates during developmental transitions (Rutter 2022, 2024; Rutter et al. 2025; Holt 2023, 2024; Rogers and Ashworth 2024). This does not imply any one pathway accounts for CAPVA in all cases; rather, it underlines the need for proportionate assessment and individually tailored, relationally informed responses. This aligns with broader arguments that children who use family violence are still children, and responses should prioritise safety alongside support, without relying on adult perpetrator paradigms (Condry and Miles 2021; Hackett et al. 2006).

If routine domestic abuse/perpetrator frameworks are not appropriate, practice alternatives must be explicit and accessible. Trauma- and violence-informed approaches may support safety and engagement when they are used to contextualise behaviour and reduce escalation, rather than to minimise harm. International participants also highlighted existing safeguarding pathways are often insufficient for CAPVA, reinforcing the need for clearly articulated, child-centred pathways which integrate assessment, de-escalation, and tailored family support.

4.4 Family Experiences

Whilst CAPVA is thought to impact around 22% of the population, researchers posit this is a ‘gross underestimate’ (O’Hara et al. 2017, 181). Thus, many practitioners could be working with families experiencing CAPVA without them disclosing it. Removing the burden on parents to make a full disclosure which has been identified as a significant barrier to support in other violence research (Francis et al. 2017). CAPVA has many identifiable features beyond observable physical harm from children to parents; and viewing CAPVA as a collection of many different forms of harm, may be useful in terms of other forms of harm being potential indicators of CAPVA too (Holt 2025).

Parent blame is one of the most challenging experiences for families when they seek support for their experiences of CAPVA (Condry and Miles 2014; Coogan 2011; Rutter 2022, 2024; Rutter et al. 2025; Holt 2016; Miles and Condry 2015). Whilst the respective Delphi studies critique this blame, highlighting it not only prevents families from seeking support, but also undermines the relational practice which is vital in working with families experiencing CAPVA. CAPVA is not a phenomenon which benefits from attributing blame, but benefits from understanding individual and collective agency and vulnerability, as well as the complex ecological interactions which can increase vulnerability to experience or instigate harm (Baker et al. 2025).

4.5 Practitioner Work

The absence of statutory duties means many services and localities do not provide specialist CAPVA intervention. Furthermore, there is currently no statutory guidance within UK policy regarding the components of CAPVA specialism (Holt and Lewis 2021). This gap is further accentuated when CAPVA presents in early childhood, with harms reported in children as young as 3 years (Rutter et al. 2025). This can often result in families not knowing where to get support and then turning to police when the harm escalates over time, there becomes a risk to other children in the home, and/or a risk of child removal (Rutter et al. 2025; Condry and Miles 2014; Holt 2025; Holt and Lewis 2021).

The importance that multi-agency working does not rely solely on parent self-referral (thereby increasing the burden of disclosure) is an important consideration in CAPVA (Francis et al. 2017). The Delphi 2 advisory group outlined how multi-agency strategies such as Operation Encompass (Jobson et al. 2025) and Operation Soteria (Stanko and Crew 2022) were perceived as effective examples of coordinated, victim-focused system responses. Within this locality, the advisory group noted a substantial proportion of children discussed in practice were on a neurodevelopmental assessment pathway (‘neuropathway’). This observation should be interpreted cautiously and contextually: it may reflect local referral patterns and service configurations rather than indicating CAPVA is primarily neurodevelopmental in origin. Nonetheless, it usefully illustrates a service-access challenge. The advisory group highlighted where families pursue assessment to access support, they may experience diagnostic overshadowing (i.e., where a potential or confirmed diagnosis displaces attention to co-occurring needs, including harmful behaviour). Consistent with the Delphi 2 statements, participants argued support should address safety and behaviour alongside assessment processes, rather than waiting for diagnostic outcomes (Molloy et al. 2023; Rutter et al. 2025).

4.6 Peer Support

Peer support was valued across the studies as a mechanism for reducing isolation and enabling shared learning; however, it was not positioned as a stand-alone solution. The findings support blended provision (peer, one-to-one, and whole-family work) selected according to family circumstances and risk, with appropriate safeguarding, supervision, and support for peer contributors.

5 Limitations

Several limitations should be noted, which also indicate priorities for future research. First, some Delphi 2 statements were multi-component, potentially reducing item clarity and lowering interpretability of ratings. Second, Delphi 2 was a small panel drawn from a single locality and service system; the resulting consensus may therefore reflect local practice norms, referral architectures, and resource constraints. Future research should replicate this Delphi exercise across multiple local authorities and services to examine which principles are robust across contexts and which are system-contingent.

Third, Delphi 1 deliberately combined experts-by-experience and expert practitioners; this enhances participatory legitimacy but may also mask meaningful differences in priorities. An analysis of different responses between these groups is available (see Rutter 2021), however future research should be designed to enable subgroup analyses (lived experience vs. practitioner; sector; country; role) and to explore where disagreement reflects divergent values, risk thresholds, or system constraints. Finally, this manuscript does not include the perspectives of children and young people, which are central to understanding CAPVA and to assessing the acceptability of proposed responses. Future research should incorporate children's and young people's voices using developmentally appropriate, trauma- and safeguarding-informed methods, and should move beyond consensus to outcome-focused evaluation.

6 Resource Needs and Implementation Considerations

The applied nature of CAPVA work means resource implications are integral to interpretation. Across both Delphi studies, participants emphasised the need for workforce development (universal and specialist CAPVA training), distinct and navigable referral pathways, and sustained relational support rather than time-limited contact. Delphi 2 also indicated the importance of interim support during waiting periods, clearer information for families at the point of referral, and strengthened multi-agency coordination (including with neurodevelopmental and mental health pathways). Where peer support is developed, implementation should account for safeguarding, supervision, and the risk of over-burdening parents with lived experience, while retaining the potential benefits of shared learning and stigma reduction.

7 Conclusion

These two Delphi studies identify areas of consensus regarding principles for understanding and responding to CAPVA. Across both international experts and local practitioners, CAPVA was consistently framed as a public health concern and a family support issue that requires timely, developmentally informed responses. Importantly, the findings should be interpreted in light of heterogeneity: CAPVA occurs in diverse family contexts and may be associated with multiple, overlapping pathways and needs. Accordingly, children who use harmful behaviour should be understood as children first, and responses should prioritise safety alongside support rather than relying on adult perpetrator paradigms.

Participants cautioned against the routine application of traditional domestic abuse frameworks to CAPVA, highlighting risks of criminalisation, stigma, and ineffective or misaligned support. In their place, the consensus statements point toward the development of flexible, principle-led CAPVA pathways which can be adapted to local service systems. These include holistic and proportionate assessment, attention to family dynamics, and longer-term relational work where required. Both studies also identified systemic gaps which included limited awareness, unclear thresholds and routes to support, and challenges associated with diagnostic processes which can delay help-seeking and timely intervention.

Practitioners emphasised the importance of a public health approach, clearer referral routes, and multi-agency collaboration reduces the burden on parents to disclose in crisis. They also highlighted workforce development needs, including both universal awareness-raising and more specialist CAPVA training to support identification, assessment, and response. Peer support was consistently valued but was positioned as one component within a wider suite of options (e.g., group, one-to-one, and whole-family work), to be selected and combined according to family circumstances, developmental stage, and risk.

Overall, the findings support the need for coordinated guidance, improved professional training, and strengthened early support systems, while recognising CAPVA presentations and service needs vary. Service development should therefore prioritise flexible, tiered pathways that are child-centreed, non-stigmatising, and proportionate, with capacity for multi-agency working and sustained relational support where required. Future research should test the feasibility and outcomes of the proposed principles in broader, multi-site samples and, critically, incorporate children's and young people's perspectives to strengthen the evidence base for acceptable and effective responses.

7.1 Implications for Research

Participatory research with families and practitioners was effective in the development of these two Delphi studies. They also brought to light the significant differences and similarities between the two groups in how they understand and respond to CAPVA. Further participatory research, particularly exploring how parents make-meaning of these service-based interactions, and the different parenting strategies they apply, may assist in the creation of new knowledge and understanding.

7.2 Implications for Practice

Operationally, the consensus supports clearer thresholds and shared decision-making across statutory partners (children's social care, education, health, policing), with agreed routes for information-sharing and escalation when risk is acute. It also cautions against routine adult perpetrator frameworks that can undermine engagement and obscure support needs. Given heterogeneous CAPVA presentations, responses should be proportionate, developmentally informed, and embedded in tiered pathways offering rapid de-escalation and safety planning alongside routes into specialist, longer-term support, without relying on parent persistence or repeated crises.

1. Prioritise early engagement with families is non-judgemental and relationally focused, alongside proportionate assessment of safety and support needs, so responses can be tailored to the family's circumstances.
2. Embed consistent, non-stigmatising communication and co-production with families, supported by multi-agency working and clear pathways offering a tiered range of options (advice, brief support, specialist intervention) rather than a one-size-fits-all model.
3. Plan and deliver support on the basis CAPVA may be associated with multiple, overlapping pathways (e.g., trauma/adversity, neurodevelopmental difference and sensory needs, mental health, family stressors). Avoid diagnostic overshadowing by addressing harmful behavior, safety, and family support needs alongside (not only after) assessment processes.
4. Develop and/or strengthen statutory and local guidance clarifying which services hold responsibility for CAPVA responses, including thresholds, referral routes, and information-sharing, while enabling proportionate, flexible provision matched to local service configurations.