The Feasibility and Acceptability of an Experience-Based Co-Design Approach to Reducing Domestic Abuse

Abstract

One means of reducing violence against people experiencing domestic abuse is to improve the pathway which they use to access help from the police and other services. In this paper we report and reflect on a project which contributes to violence reduction via a participatory approach to service improvement, focusing on this pathway. We describe the four phases of an innovative experience-based co-design (EBCD) project, which involved collaborating with domestic abuse survivors as well as members of the police and domestic abuse organizations. We report on indicators of the acceptability and feasibility of EBCD in this context. We also reflect upon the potential of the EBCD approach for involving communities in collaborating with services to reduce domestic abuse. We discuss the conceptual and methodological implications with regard to adopting participatory and inclusive approaches in contexts where power-sharing may be difficult. We argue that EBCD has considerable potential for use in this setting and we identify several areas where insights from this project could be used to improve the future viability of any such initiatives.

1. Introduction

The World Health Organization estimates that 1 in 3 women globally are affected by domestic abuse in their lifetimes [1]. Domestic abuse-related crimes include the occurrence of threatening behavior, violence, or abuse (physical, psychological, emotional, sexual, financial) between intimate partners or family members, who are aged 16 years and over, regardless of gender or sexuality [2]. The Crime Survey for England and Wales ending March 2020 evaluated that 5.5% of adults aged 16 to 74 years (2.3 million) experienced domestic abuse in the previous 12 months [3], which is an estimated 7.3% of women and 3.6% of men [4]. In the year ending March 2021, 18% of all offences recorded by the police in England and Wales were domestic abuse-related crimes and prevalence is increasing, rising 6% from the prior year. [3] Domestic violence is not typically an isolated event: case study analyses consistently show recurrent patterns of abusive behavior (e.g., see Katerndahl et al. [5]).

One way to reduce violence against people experiencing domestic abuse is to improve the acceptability and effectiveness of the help-seeking pathway. Prevailing models of help-seeking tend to characterize the problem as an issue of individual knowledge, self-appraisal, and reasoned action [6]. Domestic abuse is an important context here because the problem and pathway are complex. The problem is complex because identifying and accepting that abuse is taking place, in the context of a familial or intimate relationship, can be difficult. Deciding to seek help may be shaped by a range of cultural, social, and relational factors [7], and by the opportunity, knowledge, and access required in order to approach someone who can help [8]. The pathway is also complex because it is provided by a range of organizations, with different interests (e.g., support and safety; housing and security; law and order; risk and child welfare), whose collaborative arrangements are varied, and opaque to the help-seeker. Some of these organizations have limited or unstable funding, and some may have different ‘thresholds’ for providing help. The help which a person hopes for may not be the help that they are most likely to receive

A report by a UK-based domestic abuse charity highlighted that many survivors went to the police multiple times before obtaining effective help [9]. It is crucial that domestic abuse survivors receive help which they find effective when they first seek it, because at the point when a survivor obtains help, the abuse may be escalating in either severity, frequency, or both [9]. This can have serious and negative implications for their safety [9]. Furthermore, survivors who are satisfied with police services and find their services helpful are more likely to contact the police again, if needed [10]. Therefore, it is important to understand what survivors find helpful and unhelpful when seeking help and to design services to meet these needs. One way of achieving this is by involving domestic abuse survivors in service development. This would ensure that services are tailored to domestic abuse survivors’ views and real needs and therefore improve the help-seeking pathway both for survivors and service providers.

This paper reports and reflects on an attempt to contribute to violence reduction via a collaborative community-based approach to service improvement, called experience-based co-design (EBCD [11]). EBCD is a participatory, action research process, which was originally developed as a tool for improving patient and staff experiences of healthcare services [12]. Co-design is a form of community-based action which involves working closely with stakeholder groups to make shared decisions about how to improve a common resource, important process, or shared environment. EBCD is a relatively formalized approach to co-design, with accessible steps and strategies, which can be implemented in public services. As a result, it is becoming an increasingly important tool in the development of such services [13,14]. Here we aim to contribute to the field of violence reduction by reporting on a novel implementation of EBCD in this context, and to the development of EBCD itself, by discussing the approach in the context of a conceptual framework which helps us to consider how EBCD work shifts the traditional relationship between ‘service provider’ and ‘service user’.

EBCD has been used in a range of community contexts (physical healthcare, mental healthcare, learning disability, interventions research) to collaboratively create a wide range of ‘things’—information resources, service improvements, built environments, implementation pathways (e.g., see Dimopoulos-Bick et al.’s synthesis [15], and Donetto et al.’s review [13]). For example, a survey has found that EBCD projects have been undertaken in the following range of clinical services: cancer, diabetes, genetics, drug and alcohol services, intensive care, emergency services, palliative care, orthopedics, surgical units, hematology, and neonatal and pediatric care [13]. The findings of the survey have also demonstrated that at the time it was conducted, EBCD projects were either achieved or being conducted in the following countries: UK, Canada, New Zealand, Australia, Sweden, and the Netherlands [13].

EBCD has not previously been implemented in the context of policing and domestic abuse and so our primary purpose in this paper is to describe the process that we undertook, and to reflect upon the acceptability and feasibility [16] of this approach for community-based approaches which involve a range of partner-organizations. Thus, the project described here increases our knowledge of how EBCD works, for whom, and in what contexts, especially surrounding the issues involving domestic abuse. This paper aims to answers the research question, ‘Is it feasible and acceptable to conduct EBCD in the field of domestic violence?’.

Our project took place in a major conurbation in England. The conurbation is ethnically diverse. It has been undergoing a long transition from an economy based on heavy industry to a much more mixed economy. It was jointly commissioned by a Police and Crime Commissioners Office and a local police force. The project was initially commissioned to explore satisfaction of domestic abuse survivors with their services and to use these findings to provide recommendations for service improvements. The local police force approached researchers at the University of Birmingham in 2015 with proposals for a survey of domestic abuse survivors. Through discussion with representatives of the police force, it was agreed instead to use an EBCD framework, first to understand the experiences of survivors, and then to collaboratively plan service developments. As partial funding came from the local police force, the study was based in the region covered by its service.

EBCD begins with a phase of finding out about people’s experience of a particular process or environment. These insights are organized as ‘touchpoints’—features of an environment or process that make a difference to people’s experience. EBCD then moves to a ‘feedback’ phase, which involves consulting with the different stakeholder groups to discuss the touchpoints and generate consensus about what needs to change. These stakeholder groups may include patients, staff, and carers. It concludes with a co-design phase, where the different stakeholders work together to decide how to make improvements, and who should make them. In this way, the different stakeholders in a given community work together to identify things which could be improved, and they continue to collaborate in order to decide how to make those improvements.

An EBCD approach prompts us to reflect upon help-seeking as a systemic and relational activity precisely because it brings together different stakeholders. The involvement of survivors from different communities prompts us to consider accessibility and knowledge. The involvement of different elements of police services (e.g., specialist and generic; call handling and first response) prompts us to return to the question of knowledge from a service perspective, and to consider it in the context of communication with help-seekers, and the consistency and empathy of the response. The involvement of third-sector services brings the issue of communication between services into view. When we consider these different groups as members of a community of people who are collectively ‘responding to domestic abuse’, we see a shared value: domestic abuse is wrong, and something should be done about it. However, we also see a range of different views about which actions to prioritize in responding to that.

It is certainly possible to imagine a community-based EBCD project which does not involve the police as key partners, but in this project, the police played an important role. In many ways, this was positive—the nature of policing organizations is such that they have high standards for governance, internal organization, and project management. These assisted with recruitment and engagement to the project, and with subsequent commitments to implementing the project’s recommendations. However, it is also important to consider the implications of police involvement upon a method which is committed to collaboration and community-led change.

Epistemic justice [17] is a conceptual framework which is helpful in this regard, and which can help us to think more generally about the dynamics of coproduction—and particularly some of the complexities involved in implementing a community-based approach when some community-members (i.e., perpetrators) are deliberately excluded from the work [18,19] and where other community-members (i.e., police) hold considerably more power than the vulnerable population at the center of the work [13]. This conceptual framework approaches the question of knowledge from a philosophical perspective which is concerned with equality and justice. For example, knowledge about what to call a problem, how to identify it, and what to do when you encounter it. There are some direct and immediately helpful entailments, in terms of the way that Fricker’s terminology conceptualizes the disadvantages and injustices experienced by those survivors of violence who are—for example—situated in social and cultural environments which do not consider intimate partner violence to be unacceptable, or which grant spousal perpetrators certain exceptions, or which understand help-seeking from outsiders to be shameful. These are all forms of hermeneutic injustice, because they involve a person who is disadvantaged by knowledge that they do not have or cannot access. Similarly, Fricker’s concept of testimonial injustice applies to the many situations in which survivors of violence are not believed due to the biases of the person from whom they seek help: they may not be believed by in-laws, for example, because the perpetrators are family members; or by the police, because the survivors are either from communities where domestic abuse is not expected, or conversely, is not problematized by the police. Therefore, our secondary purpose in this paper is to reflect on some of the workings of a co-design approach to domestic abuse through the lens of epistemic justice, and a means of considering its potential contribution to harm reduction. This framework was not used to plan our implementation of the study. We followed the standard pragmatic process of EBCD. However, we have found it useful as a means of reflecting on what happens during the implementation of EBCD work, and of linking to established models of the acceptability of interventions [20]. This paper therefore outlines the methods used in each stage of the project with accompanying reflections on the acceptability and feasibility of those methods, followed by a discussion on the overall process and potential impact of the study, drawing on the existing literature.

2. Our Project

The project started early in 2016 and was guided by a steering group, gathered together by the local police force, with guidance from the research group, and including representation from police. It was usually comprised of two police officers, one of them the senior officer acting as project lead from within the police; two domestic abuse survivors; the lead for local authority’s co-ordinating domestic abuse organization, and the research team (n = 2–3). The group met once every two to three months for the lifespan of the project and meetings were usually chaired by the senior police officer.

The group set and followed an agenda, but the chair took care to actively invite contributions from all the different stakeholders. As a result, many of the challenges this project faced were resolved through this group. Thus, when the research team felt ‘stuck’ with something, the other members of the steering group would furnish them with a new set of strategies. For example, this occurred when trying to recruit participants for the project. Furthermore, the insights of the survivors, police officers, and support lead on the steering group were invaluable. For example, their help was especially important when it came to designing the interview schedules, developing recruitment strategies, reviewing the touchpoints from the research phase, and assisting in planning and running the co-design event. Consequently, this had a positive impact on the outcome of the research project. For example, they greatly helped in recruiting participants and ensured that the interview schedules were sensitive to participants’ needs.

The EBCD project was conducted in four phases, described in detail in the following sections. See Figure 1 for an overview of the process. The first two phases focused on stakeholders within one city in central England. These stakeholders included domestic abuse survivors, representatives of the police, and independent domestic violence advisors (IDVAs) who provide specialist support to survivors. To increase participation, the latter two phases expanded the range to include stakeholders from the wider region but still within acceptable travel times to allow in-person meetings. The first phase, which acted as the research phase, involved gathering experiences of providing or using services to generate a list of ‘touchpoints’ or features of this process which had an impact on how it was experienced. In this project, we decided to refer to touchpoints as keypoints, because survivors preferred it. In the second phase, in which the feedback groups were conducted, we brought together groups of stakeholders to further discuss the touchpoints and to prioritize areas for action. The third phase involved the co-design event where stakeholders came together and engaged in solution-focused discussion to generate action plans. In the final stage of the project, which acted as the implementation phase, the police were responsible for observing and implementing these action plans. The figure below describes the phases of the project, their aims, and who was involved. It is here to set the scene for the detailed description which will follow and thus enhances our understanding of how the project advanced from one stage to the other.

2.1. The Research Phase

• Recruitment

The aim of the research phase was to gather the experiences of domestic abuse survivors, the police, and the IDVAs. We set out to recruit service providers (which included representation from the police and IDVAs) and domestic abuse survivors, who could provide personal experiential perspectives on the pathway for accessing and receiving help. Domestic abuse survivors were eligible to participate if they had experienced domestic abuse within the past 24 months. We aimed to include a diverse range of perspectives. We used purposive sampling to pursue this. In the case of service providers, purposive sampling allowed us to ensure that we included participants with a wide range of roles. In the case of service users, we included participants with a range of ethnic backgrounds, and with a range of differing relationships to the perpetrator.

Domestic abuse survivors were recruited via gatekeepers from domestic abuse support organizations and the police force in the city. Service providers were eligible to participate if they were involved as a professional in a service provided by a support organization or the police force of the county this project was conducted for. The service providers were recruited with the assistance of the members of the steering group, who were involved in the services provided by either the police or the domestic abuse organizations.

• Data collection

We conducted either semi-structured face-to-face individual interviews or focus group discussions, in English, with our participants. Most of them were carried out by the authors, but a small number were carried out with support from master’s students. Both the interviews and the focus groups were open-ended data collection events, facilitated by an interviewer drawing on a set of exploratory questions in a topic guide. The questions in the topic guide were generated by consulting with representatives of all stakeholders at the steering group. Thus, we prepared schedules of open-ended questions to explore participants’ experiences of the help-seeking process, and we employed these flexibly, following established standards for semi-structured approaches to qualitative data collection [21].

Data collection lasted 14 months; the service users were interviewed during an eight-month period within that window. We continued to try to recruit service users for the remaining period of four months but were unable to increase our sample size. It proved to be challenging to recruit survivors for the project due to the vulnerability of the survivors and potentially changes in the circumstances of survivors who showed initial interest in participating.

The interviews lasted 41–121 min. Focus groups lasted approximately two hours. All of the interviews and focus groups were audio-recorded in full and transcribed verbatim.

• Participants

Six survivors (five women; one man), participated in this phase of the project. All the survivors were parents, had received services from an organization which supports domestic abuse survivors, and had suffered from multiple incidents of abuse. The abuse had occurred within a context of a heterosexual relationship.

Twenty-two police participants took part across ten individual interviews and three focus groups, including seven senior police officers. They included: initial response police officers, non-urgent and urgent call handlers, force contact management, dispatch and resource allocation staff, offender management, specialist police officers working for the Public Protection Unit who investigate domestic abuse offences, and those involved in policy and strategy related to domestic abuse.

Three IDVAs, all women, participated across one individual interview and one group discussion. One IDVA was a court IDVA. A fourth, male participant, was a service delivery officer. For the purposes of brevity, we will refer to them as IDVAs.

Interpretation of data in EBCD is often conducted relatively informally [11] but there are also some advantages to taking a more formalized and systematic approach [22]. In this project, the authors analyzed full transcripts of the interviews. We identified keypoints using an inductive, open coding strategy in the manner of the early stages of a thematic analysis. From the transcripts, we systematically extracted each claim that was made about the valence of an experience (e.g., whether it was good or bad), alongside a code about what this meant to the participant, and then recorded all of this information in a coding framework. Rather than developing ‘themes’ per se, after reviewing all of the extracted keypoints, we sorted them into groups based on their shared concerns and translated each group into a single paraphrased statement. The identified keypoints are listed in

Table 1. List of the keypoints drawn from the accounts of the participants ¹.

Keypoints Drawn from the Accounts of the Survivors, Police Employees, and IDVAs

Services need to be ready to support people who will have different ‘tipping points’ for seeking help. Services need to be ready to support people who have had ‘help’ imposed upon them. Services need to be ready to support people who want different sorts of solutions. The police are perceived to be ineffective, or even make things worse. The police will not understand. It can be difficult to get/find help. All professionals need to be aware that children, or the relationship with the children, may be put at risk. Organizations can leave survivors feeling excluded from their processes. Positive contact can help survivors to feel safe and reassured. Once other people were involved, their support/involvement was often very helpful to survivors. It is perceived to be difficult to get effective help when your problem does not fit the mold. It is vital for the police to get good information in order to tailor their response. A good response is a prompt and effective response. Police work is psychologically taxing—relatively informal resources are used to support this. Some practices can undermine people’s willingness to return to the police for help. Organizations can be poor at communicating with each other. There are tools which are not always used or are not available everywhere or are not effective. It can be difficult for the police to know how to make a difference in domestic abuse cases. It is important to educate the public about abusive behavior.

¹We do not provide a detailed presentation of the findings from the research phase of this project in this paper, because our aim here is to give an overview of the approach that we took, and to share our insights with regard to its acceptability and feasibility for harm reduction approaches directly involving communities.

, below.

The findings from the research phase highlighted that domestic abuse survivors could find it difficult to get/find support that was sensitive to their needs from the police and domestic abuse organizations. Their negative experiences of help-seeking were seen as undermining their willingness to return to the police for help. The police recognized that their services were variable and that unfortunately not all survivors received a good response and therefore that poor services do need to improve. The police and IDVA participants acknowledged that survivors might want cases to be resolved in different ways (e.g., some survivors might want the perpetrator locked up, while others might want to stay in the relationship but to have the abuse stop). Another important area of focus of discussion for all three stakeholder groups was around the importance of communication. For example, the domestic abuse survivors highlighted that it was important to them that there was good communication between them and their formal service providers. Similarly, service providers spoke about the importance of good communication in their work. The IDVAs were frustrated that although they shared information with the police, the police did not always do the same with them.

In EBCD, the identification of touchpoints/keypoints is an important step in providing focus to the co-design work. The keypoints act as a stimulus for further conversations about what should change, and how it should change, during the remaining phases. Crucially, this still leaves space for the inclusion of new perspectives (in our case, both additional survivors, and wider professional perspectives), and further refinement of the co-design community’s aims.

Interim Discussion: Indicators of Acceptability and Feasibility during Recruitment and Engagement

It is important to reflect on the acceptability and feasibility of this phase of the project for potential further initiatives. It was certainly feasible to recruit service providers to this project: both police and IDVAs were quick to understand the project, and soon engaged with the research phase. There was often a good initial response from survivors too. Many expressed an interest in taking part, but when it came to arranging interviews, there were often difficulties in making contact with survivors, in setting up suitable interview arrangements, or in survivors attending the interview. We were not able to collect substantive feedback on why these difficulties arose, but through discussion with various stakeholders on the steering group and beyond, we tried a range of alternative strategies to invite and involve survivors as research participants. A further challenge stemmed from some survivors being in too vulnerable a situation to participate (e.g., due to concerns that the perpetrator would find out). Allowing for eligibility of service users to participate for a longer period after the abuse had occurred would have helped with this. Services themselves had little time to support recruitment of participants, but there was better engagement when they were able to do this. From the perspective of the epistemic justice framework, these issues underline the power differentials which are in play during the opening round of a co-design project. Projects may reproduce or amplify hermeneutic injustices, because service users do not know what the process will involve, or what will be asked of them, prompting wariness about getting involved. If their prior experience of the services has been poor, it may fuel expectations of testimonial injustice—that they will not be believed. Trust is clearly crucial in this context.

The data collected also had several limitations. First, four of the domestic abuse survivors agreed to be interviewed in English but were not as fluent in English as they were in their first languages. The project did not include funding for interpreters, but these participants may have been able to express themselves more comfortably, and more eloquently, if we had been able to use interpreters. Second, all the survivors participating in the first phase of the project had sought help, lived in the same region of the UK, and had children. The sample only included one male participant. We tried to compensate for this by including the perspectives of service providers who described the needs of other male survivors. This does mean that our starting sample did not account for the full diversity of survivors of domestic abuse. However, the data collected had two key strengths. First, they successfully captured the participants’ lived experiences of help-seeking/providing support, as well as the impact these experiences had on them. Second, for a qualitative study, where the focus typically prioritizes depth over breadth [21], it included a relatively large sample of service providers.

2.2. The Feedback Group Phase

In EBCD, the primary purpose of these groups is to discuss, cluster, and prioritize the keypoints, in order that key areas of consensus can be taken forward to the co-design stage. An important secondary function is that these groups support perspective-taking: they allow stakeholders to see what other groups have said, and to prepare for the collaborative work of the co-design event. The groups also widen the range of perspectives involved in the process, expanding out beyond the initial interview sample. We conducted 12 feedback groups with approximately 40 participants. At the end of each feedback event, the outcome was a list of the clustered and prioritized keypoints, as these were arranged by the stakeholders. We recorded these priorities in a spreadsheet and mapped the areas of agreement across the feedback events.

2.2.1. Results of the Feedback Groups

This produced five priority areas of consensus: (1) Having an open mind about who needs help and being ready to provide a humane first contact; (2) A range of options for responding which do not place further burdens on the survivor or their children; (3) Developing support and training for police officers; (4) Improving knowledge about when and where to seek help, and what to expect; (5) Improving information sharing and collaboration across organizations. The first four areas of consensus were identified by all stakeholder groups. In other words, they were identified by the survivors, police, and support organizations. The fifth area was primarily prioritized by the police and the support organizations but was also evidenced by the survivors.

2.2.2. Interim Discussion: Indicators of Acceptability and Feasibility during Feedback Events

During this stage it was comparatively easy to recruit participants from all stakeholder groups. We began to widen our geographical range, from city to region, for some of the feedback events. However, for survivors it may well also have helped that in this phase they could attend events together. This underlines the relational context of epistemic justice: it is therefore important in planning co-design to consider the dynamics involved in co-design activities. Setting up events with a collective service user presence may reduce the perceived threat of testimonial injustices. This observation is reflected in the wider coproduction literature. In addition, participants did not have to talk about their personal experiences to take part in these events, further reducing the potential inequities or anticipated challenges which might have discouraged participation in the previous phase.

Participating in this stage of the project appeared to be generally easier and more acceptable for survivors. On reflection, a collective means of participating in the research phase, rather than individual interviews, may have been helpful for engagement. However, our priority in adopting interviews for the first phase was to use a medium which would allow us to provide sensitive responses to any disclosures of risk or distress.

2.3. The Co-Design Event

The third phase of the co-design project was the co-design event. The co-design event took place at another university in the host city. An academic setting was chosen partly because it represented ‘neutral ground,’ and partly because the police wanted to highlight the strong academic emphasis that had been put into the project and the fact that it was conducting innovative work. The research team took the lead in organizing the co-design event but consulted with the other members of the steering group to do so effectively. For example, the members of the steering group decided together on the venue, as well as on whom to invite, but the research team hired the venue and sent out the invitations. In this project, the co-design event was very important since it brought all the stakeholders together to design an action plan to improve services based on the consensus areas. Forty people attended the co-design event. It developed around five working groups, with each of them focusing on one of the different consensus areas. Each group was assigned a facilitator, to keep the group ‘on task’, and a ‘champion’ to provide organizational advice on how plans could be implemented, and by whom. The facilitators were chosen by the steering group as those they personally knew from work and considered to have the best skills to fulfil this role. Furthermore, the champions were chosen by the steering group because they were people with organizational roles which would enable them to bring the action plans forward due to their senior positions and responsibilities. The aim of the co-design event was to create a space in which domestic abuse survivors and service providers could work collaboratively to design action plans to improve services. Therefore, mixed groups were created around each of the consensus areas. To support the action planning, groups were provided with a simple template, prompting them to record their aim, the steps needed to execute their plan, and an evaluation plan (how they would know when they had achieved it). Following this work, the groups were invited to share their plans in a series of short presentations.

2.3.1. Results of the Co-Design Event

Each of the five working groups produced at least one action plan. Some groups produced more than one plan, and some groups coordinated their plans to complement the work of other groups. A total of seven plans were proposed. These are briefly reviewed below in

Table 2. Proposed action plans according to working group.

Working Group	Proposed Action Plan
Group 1: Having an open mind about who needs help and being ready to provide a humane first contact	Proposal to improve survivor experience by providing more specialist support, with trial of domestic abuse specialist car (as per mental health triage team; police officer to be accompanied by peer support worker or IDVA) to provide timely follow-up to the first response.
Group 2: A range of options for responding which do not place further burdens on the survivor or their children	This group proposed two action plans: To reduce risk to survivor by sharpening focus on perpetrators through more routine discussions and monitoring of perpetrators at multi-agency meetings. To give safe period of reflection, post-first response, to the survivor, by development and trial of ‘brief stay’ respite accommodation for perpetrators, staffed with specialist worker.
Group 3: Developing support and training for police officers	Proposal to reduce variability of first responses by providing mandatory training to all response officers; this training would be delivered by a range of media (including face-to-face and survivor-led) to improve their understanding of the complexities underpinning domestic abuse, and the range of appropriate responses available to them.
Group 4: Improving knowledge about when and where to seek help, and what to expect	This group proposed two action plans: To improve help-seeking in the longer-term by improving young people’s knowledge about healthy relationships with a campaign/education program in schools. For a survivor-informed review of currently available information (about domestic abuse and the help which is on offer) to identify areas for improvement.
Group 5: Improving information sharing and collaboration across organizations	Proposal to improve access to support by mapping the available services and tools, and then developing an online resource that provides details of different pathways and the help available at different points on those pathways, produced to link directly to relevant agency websites.

, according to working group.

The underlying priority areas provide useful context for thinking about how these plans might be implemented, extended, given a sharper focus, or supplemented by further initiatives. For example, the group briefed to develop support and training for police officers focused on trying to improve consistency of first responses by proposing mandatory training on domestic abuse. However, they also had support in their remit, and there was considerable discussion during the feedback groups about the pressures of staff wellbeing, and the lack of available support after exposure to trauma or stress. Although this group found staff well-being an important topic, they did not design any action plans for this, but only created action plans for the training component. In retrospect, it might have been better to separate this group into two with one group focusing on staff training and the other on staff well-being.

2.3.2. Interim Discussion: Indicators of Acceptability and Feasibility during Co-Design Event

During the event we observed collaboration and communication between the different stakeholders, both within and between the working groups. The impact of survivor testimony was powerful, and clearly influenced the discussion and direction of each group’s work.

At the end of the co-design event, feedback from the participants was gathered through an open-ended evaluation form. Although the form was anonymous, the participants could indicate if the feedback came from a survivor, family member, police employee, support professional, or another role. They could tick more than one option (e.g., if they were a survivor who also had a professional role in a support organization). Twenty-three of the forty participants completed the feedback form. Overall, the feedback from the event was excellent. The participants indicated that the presentations, which were held during the event, were interesting, easy to follow, and respectful of everyone’s point-of-view. Furthermore, the groupwork also received positive comments with participants indicating that: it was easy to understand what they had to do; the group leaders kept everyone involved and on track; it was respectful of everyone’s point-of-view; and useful since they came up with a plan that could make a difference. Moreover, the responses of the service providers indicated that: obtaining the feedback of domestic abuse survivors gave them a better insight; it helped pinpoint areas for improvement; they would like to take what they learned into their work; it was productive to have different perspectives at the table; the working groups were balanced in terms of stakeholders; and they hoped that the ideas that they generated would become a reality. On a less positive note, the feedback from some of the service providers also indicated that although the co-design event generated good ideas it was difficult to formulate the steps needed to execute them and there was not enough time to do so during the event either (the groups had an hour and 40 min to do so, which included a ten-minute break). The feedback from the survivors specified that they appreciated hearing the views of police employees, especially those of call handlers and initial response officers, and that they hoped that it would make a difference in the field of domestic abuse. By this stage, the co-design process appeared to be highly acceptable to stakeholders from all groups.

The steering group met once following the co-design event to discuss their observations from the event, the feedback that was obtained, and the handover of the final report. Following this meeting, the research team wrote a short report about the EBCD project for the police. The interim report summarized the process and recommendations from the co-design component. For example, it included the procedure of the whole project, the identified keypoints, results of the co-design event, and illustrative anonymized quotes of the participants. The report also included the feedback of the domestic abuse survivors who were part of the steering group about the process of being involved in an EBCD project. They provided a personal statement, which is included in

Table 3. What is it like to be a survivor involved with an EBCD project?

What Is It Like to Be a Survivor Involved with an EBCD Project?
“As a survivor I have found being part of this research a huge success. I have been able to share my story of domestic abuse and how it has affected myself and my children, highlighting things that went well and things that went wrong. I have been able to rebuild positive relationships with many people working in the very service I had felt let down by. This process has been hard at times—opening up about difficult experiences. However, I believe this has contributed massively to my healing journey in a positive way. I have become much more confident at public speaking and this is helping me a great deal in my [title] degree”.
“Being part of this research project has played a major role in my healing and recovery. Just knowing that it will benefit other survivors/victims, throughout different organizations, made it worthwhile and something that I not only wanted to do but felt I needed to do. I experienced some truly shocking responses from different organizations that were meant to help, so the chance to try and correct that for others wasn’t one to be missed. I am really proud of the work and the dedication from the team. I hope that with it we can make a difference and others will receive the correct help that I so desperately needed”.

, below.

These powerful comments from survivors who were involved with the project [23] show how a sense of shared enterprise may arise from being involved in a community-based participatory approach. Here, there is also a sense that participation in the co-design process may also contribute to people’s own recovery journeys. These comments underscore the moral value of helping to make a difference, and how that comes with some personal cost, but they also show why it is so important that organizations act upon the outcomes of such processes.

2.4. Implementation

In this project, funding and research governance did not extend to the implementation stage. The police took responsibility for monitoring and carrying out the action plans generated from the co-design event. The action plans were subsequently included in the Regional Strategic Plan for Domestic Violence. We do not have data on which plans were then implemented, or how they were perceived to improve help-seeking experiences and contribute to harm reduction. We consider that it has been useful and important to show that this kind of approach can be conducted with the police as partners, and that it is acceptable and feasible for stakeholders in the field of domestic abuse. Further work is required to enhance the sustainability of EBCD approaches, and to provide evidence of subsequent improvements to services, in this context. This work should include structured study of the implementation processes following on from co-design.

3. Discussion

3.1. Implications

Previous authors have observed that providing more training does not in itself lead to direct improvements in policing [14]. To improve responses to meet the needs of survivors, culture change and behavior change are required. This project has shown that it is acceptable and feasible to involve those with a lived experience (of domestic abuse) in community-based collaborative approaches to improving services and reducing harm. Threats to feasibility were overcome. However, the project faced some problems of acceptability and feasibility, especially when it came to recruiting domestic abuse survivors for the research phase. In the preceding sections, we have discussed some ways in which this issue could be resolved to promote greater community involvement at all phases of the project.

Taken together, our observations from the final co-design stage suggest that this critical step in the EBCD approach to community involvement was highly acceptable to participants. It seems that by this final stage of the EBCD process, further improvements had been made with regard to the epistemic inequalities and injustices which initially hampered the mutual understanding that is needed to agree upon changes appropriate for the community as a whole. In a sense, the novelty of the co-design event itself helped to flatten out some of these inequities: neither the service providers nor the service users had special expertise in co-design, and the process required them to consider how to draw on the different kinds of perspectival expertise (professional, experiential, or both) which they brought to the topic. However, it is also important to note that processes which began during the previous stages of the project had important epistemic ‘payoffs’ at this point. For example, data collected about all stakeholders’ concerns were presented, and weighted equally, to set the agenda for the event. The priorities set during the feedback group discussions were also presented; this is where stakeholders were also invited to consider what the priorities of other stakeholders might be, to help to prepare people to work together. Collective involvement (all of the main groups were well-represented) and the visible presence of testimonial evidence at the event meant that the threat of testimonial injustice was greatly diminished at this stage.

Working towards this kind of community involvement may be crucial for supporting the kinds of change which are required. In our project, police participants themselves informed us that they felt that they learned best when they had the opportunity to speak to domestic abuse survivors and perpetrators. Therefore, it may be valuable for the police and the domestic abuse organizations to conduct similar EBCD projects in the future, to be more ambitious in the ways that power is shared, and to develop the capacity to sustain and integrate these ways of working with communities.

In doing so, there is much to be learnt from the expertise of survivor-led movements in other domains, particularly mental health. As other observers have pointed out [12], it is crucial that organizations which draw on the lived experience of their service users make a commitment to acting on what they learn, in order to maintain trust and prevent iatrogenic harms. The degree to which such projects create an opportunity for improving epistemic justice relates directly to the extent to which they are effective, acceptable, and feasible for those they aim to involve, and for those whose experiences they aim to improve.

3.2. Epistemic Justice and Epistemic Capture

In the clustered keypoints which were the focus of our co-design process, participants identified a number of ways in which violence was perpetuated (or went unchecked). We can understand these in relation to the two forms of epistemic injustice. Some work focused implicitly on matters of testimonial injustice (e.g., see the focus of Working Group 1 on the need for more specialist support) and others in relation to hermeneutic injustice in both the short term (e.g., Working Group 4 on improving people’s knowledge of when and how to get help) and the long term (e.g., Working Group 4 on educating young people about healthy relational behaviors). Thus, community-based EBCD work can be a means of ‘building new rooms’ [24], in the sense of creating spaces where conversations can take place which do not simply reproduce existing inequalities. In these spaces, collaboration can potentially lead participants down new, more constructive routes. Our experience of conducting EBCD with the police as partners suggests that this is possible, but it also highlights some of the ways in which it is difficult.

Police organizations share many features with the health and social care services where EBCD evolved: internal hierarchies, structures linked to budgets, and distinct organizational cultures with preferred ways of controlling and managing change, and preferred ways of identifying and responding to problems, etc. In policing, many of these features are ‘writ large’ and so the involvement of more vulnerable partners (domestic abuse survivors) comes with the potential for ‘epistemic capture’ [25].

Epistemic capture refers to the risk that the knowledge produced with and by survivors may be co-opted by more powerful partners, and that survivors have little eventual say in how it is used and acted upon. In contrast, expert-by-experience researchers have argued persuasively that the route to more equitable and effective services lies not in bypassing survivors’ ownership of their expertise, but in empowering it [26,27,28]. To some extent, this highlights a limitation of prevailing models of acceptability [20]. Many dimensions of the acceptability construct resonate with issues we have discussed in this paper. ‘Affective attitude,’ ‘burden,’ and ‘ethicality’ from the Sekhon et al. model [20] seem particularly salient. However, there is a background assumption that interventions come from above, rather than being developed from the bottom up, and it may be that the acceptability concept needs further development to incorporate issues of power and justice, as co-design approaches to intervention become more commonplace.

Initially, this epistemic capture appeared to be the case in our project. After the co-design event, there was little dialogue about the action plans—though there was a commitment made to implementing them. However, previous EBCD researchers have written about the way that co-design processes can be a ‘trojan horse’ for culture change, promoting improvements to mutual understanding [29] even while stakeholders are ostensibly focused on action-planning. Interestingly, some initiatives which followed in the wake of our project appeared to acknowledge the importance of centering survivors’ experiences in determining policy. For example, there was a personal testimony event for survivors which in turn led to a revised policy plan from police commissioners.

Our project relied on those in power to offer opportunities to survivors to share power. Given that opportunity, survivors did design action plans alongside service providers during the co-design event. In addition, there was representation of survivors on the steering group, who were involved in the implementation of the project. For example, they provided their feedback on the interview schedule, developed recruitment strategies, reviewed the keypoints, and assisted with planning and running the co-design event. Thus, the co-design approach used in this project tried to encourage citizen participation by creating several platforms in which survivors worked in partnership with their service providers to co-design improved services. This does not flatten the underlying inequalities: service providers could simply decide not to implement the action plans generated from the project, if they wished. From an ethical point of view, it is obviously important that co-design processes lead to change. In our project, the co-design results were incorporated into the regional strategic plan, but as we have discussed, beyond the point of plans and policies, it became difficult to track any changes made. This is a challenge for co-design processes which often involve handover at the implementation stage.

There are means by which future EBCD projects could support greater power sharing and epistemic equality in this context. These include paying survivors for their contributions to the project; involving survivors in data collection, as interviewers, and analysis; ensuring that they are involved in the leadership team for the full duration through to implementation. A step up from this would be to support and sustain a community-led EBCD group, in order to maintain an ongoing cycle of EBCD-led service improvements, grounded in local experience and expertise.

3.3. Future Research Directions

The critical next step is evidence about the effectiveness of utilizing EBCD in this setting [20]. To gather this evidence, it is necessary to conduct studies which identify what changes are made through the co-designed plans, and which capture how these are achieved. Studies will then be required which can propose and test mechanisms by which those innovations might reduce violence and improve user experience. It is also clear from the CORE study in Australia [30] that EBCD work can be conducted at scale, across large organizations, and so a further step in terms of feasibility and effectives would involve exploring how parallel EBCD programs might be associated with more diffuse changes in culture and behavior.

4. Conclusions

This project has demonstrated that EBCD can be implemented in a policing setting with victims of crime—particularly with domestic abuse survivors—which can be adopted into police work. Thus, by using EBCD we have the potential to design action plans which improve police services in a manner which listens to the needs of the survivors as well of their service providers. We have also reflected on some implications for police practice as well as on the feasibility and acceptability of such initiatives. We have also suggested potential future research directions which would help examine the acceptability of using an EBCD approach in this context. Since using such an approach has the potential to improve services for domestic abuse survivors and their service providerswe hope that future EBCD projects in a policing setting will be implemented, so that police services can be genuinely co-designed. It is important that such implementations—and their potential effects on survivor well-being, staff workload, complaint reduction, and prosecutions—are tracked, evaluated, and made public. In conclusion, through an EBCD approach in a policing setting, we can potentially make a difference for the people who matter.