Trauma‐informed co‐production: Collaborating and combining expertise to improve access to primary care with women with complex needs

Abstract Introduction Health, social care, charitable and justice sectors are increasingly recognising the need for trauma‐informed services that seek to recognise signs of trauma, provide appropriate paths to recovery and ensure that services enable people rather than retraumatise. Foundational to the development of trauma‐informed services is collaboration with people with lived experience of trauma. Co‐production principles may provide a useful framework for this collaboration, due to their emphasis on lived experience, and intent to address power imbalances and promote equity. This article aims to examine trauma‐informed and co‐production principles to consider the extent to which they overlap and explore how to tailor co‐production approaches to support people who have experienced trauma. Methods Bridging Gaps is a collaboration between women who have experienced complex trauma, a charity that supports them, primary care clinicians and health researchers to improve access to trauma‐informed primary care. Using co‐production principles, we aimed to ensure that women who have experienced trauma were key decision‐makers throughout the project. Through reflective notes (n = 19), observations of meetings (n = 3), interviews with people involved in the project (n = 9) and reflective group discussions on our experiences, we share learning, successes and failures. Data analysis followed a framework approach, using trauma‐informed principles. Results Co‐production processes can require adaptation when working with people who have experienced trauma. We emphasise the need for close partnership working, flexibility and transparency around power dynamics, paying particular attention to aspects of power that are less readily visible. Sharing experiences can retrigger trauma. People conducting co‐production work need to understand trauma and how this may impact upon an individual's sense of psychological safety. Long‐term funding is vital to enable projects to have enough time for the establishment of trust and delivery of tangible results. Conclusions Co‐production principles are highly suitable when developing trauma‐informed services. Greater consideration needs to be given as to whether and how people share lived experiences, the need for safe spaces, honesty and humility, difficult dynamics between empowerment and safety and whether and when blurring boundaries may be helpful. Our findings have applicability to policy‐making, funding and service provision to enable co‐production processes to become more trauma‐informed. Public Contribution Bridging Gaps was started by a group of women who have experienced complex trauma, including addiction, homelessness, mental health problems, sexual exploitation, domestic and sexual violence and poverty, with a general practitioner (GP) who provides healthcare to this population, alongside a support worker from the charity One25, a charity that supports some of the most marginalised women in Bristol to heal and thrive. More GPs and healthcare researchers joined the group and they have been meeting fortnightly for a period of 4 years with the aim of improving access to trauma‐informed primary care. The group uses co‐production principles to work together, and we aim to ensure that women who have experienced trauma are key decision‐makers throughout our work together. This article is a summary of our learning, informed by discussion, observations and interviews with members of the group.


| INTRODUCTION
Co-production is an approach where professionals and people who may use or are affected by public services collaborate, using the experience and expertise of all to form equitable partnerships to develop services/ research and outcomes. [1][2][3] While co-production can help tackle inequities in health services by enabling those who are often excluded to help shape them, 1 there can be barriers that prevent involvement and perpetuate marginalisation. 1,4,5 There is increasing recognition of the needs of marginalised communities and the structural obstacles to involvement that they may face, 6 alongside support that might be needed by those who have experienced trauma. 7,8 Existing co-production approaches need to acknowledge and understand trauma and its potential impact on individuals, group dynamics and health inequalities. 8 The high prevalence of trauma in all sectors of society, and even higher prevalence among groups who experience health inequalities, mandates careful consideration of the impact of trauma within co-production work.
Trauma can be defined as 'an event, … or set of circumstances that is experienced by an individual as physically or emotionally harmful or life threatening'. 9 Complex trauma is prolonged and is often inflicted by an individual who should be trusted-for example, experience of child abuse or domestic violence. 10 The reduction in life expectancy for those experiencing complex trauma is well documented, 11,12 with higher rates of mental health problems, substance misuse, cardiovascular disease, diabetes, gastrointestinal disorders and cancer, amongst other conditions. 13 Complex trauma negatively impacts the ability to access healthcare services, experience of these services and ability to participate in research. 9 This compounds the existing direct physical and mental health impacts of complex trauma.
In recent years, there has been a move towards trauma-informed approaches to healthcare delivery. 9,14,15 An organisation/system that is trauma-informed: realises the widespread impact of trauma and understands potential paths for recovery; recognises the signs and symptoms of trauma in clients, families, staff and others involved within the system; and responds by fully integrating knowledge about trauma into policies, procedures and practices, and seeks to actively resist re-traumatisation. 9,p.9 Inherent to a trauma-informed approach is the explicit consideration of the cultural, historical and gender factors affecting health, and redistribution of power in decision-making. 9 Although the involvement of those with lived experience of trauma is a core aspect of this approach and some research has highlighted the potential for co-production approaches to contribute to this, 16 there is no trauma-informed framework to guide collaborative processes between people with lived experience and other stakeholders (e.g., researchers, clinicians and managers).
In this article, we describe how we used co-production principles in the collaborative development of a trauma-informed primary healthcare intervention called Bridging Gaps. We start by reviewing traumainformed and co-production principles, highlighting overlap as well as key differences. We then provide further details and aims of the Bridging Gaps project. We illustrate our co-production process to date, which involved using the principles of trauma-informed approaches as codes to analyse participant interviews. Findings are used to provide recommendations to those seeking to adopt co-production methods to work with people who have experienced trauma.

| Comparative analysis of co-production principles and trauma-informed principles
While co-production has been described as a 'vague' concept, encompassing a range of different collaborative approaches, 17 this article draws on the principles of co-production from highly cited definitional material 2,18-20 that recognises the vital role of expertise from lived experience in developing services and research. Key principles of co-production include: sharing power and decision-making; adequate resources and shared ownership; equality; diversity of voice; accessibility and reciprocity; valuing all perspectives; and an appreciation of different knowledge and skills. 2 Principles of co-production substantially overlap with trauma-informed approaches. Collaboration with marginalised and less privileged communities is frequently a feature of trauma-informed approaches, which may provide crucial opportunities for building selfefficacy, confidence, skills and worth. 7 An emphasis on changing existing power dynamics through collaboration with and empowerment of individuals with lived experience of trauma is a key aspect of traumainformed approaches. The six principles of trauma-informed approaches are listed in Table 1 and are compared to the principles underlying co-production approaches.

| Aims and research questions
The aim of Bridging Gaps is to improve access to trauma-informed primary care with women with experience of complex trauma/needs. This work is ongoing. In this article, we focus on how we used co-production approaches to develop the project and reflect on how we learnt to tailor these to support people who have experienced complex trauma. Our improvement work with general practices is reported separately. 40 In this article, we investigate the following question: • When working with people who have experienced multiple traumas, how do co-production approaches need to be developed to ensure safe, collaborative and effective working relationships?

| Development of the co-production group
As a general practitioner (GP), co-author L. P. has delivered a once-aweek outreach clinic in the drop-in centre of support charity, One25, for 5 years. The outreach clinic is an attempt to provide more accessible healthcare to a highly marginalised group within a trusted community space alongside services delivered by One25 but is unable to offer the full spectrum of mainstream primary healthcare and only operates 1 day a week. The women that One25 supports have experienced complex trauma and face numerous adverse circumstances such as addiction, mental health issues, homelessness, trafficking, domestic violence, sexual exploitation, having children removed from their care and street sex work.
All the women One25 works with have experienced trauma, and it offers specialist services alongside an ethos of nonjudgemental, unconditional love. Through her clinical work and conversations with staff and women attending the One25 drop-in centre, L. P. identified that the existing mainstream primary care system was largely not accessible to the women, despite high levels of clinical need. It was hoped that by bringing the right experts together (GPs, women with lived experience, One25 charity staff, researchers), better solutions could be developed. This led to the creation of Bridging Gaps.
Co-production meetings were held initially weekly then every 2 weeks and took place in well-known community spaces that One25 already used to provide services. Participants were offered shopping vouchers as a thank-you for their time and contributions following National Institute for Health and Care Research (NIHR) guidance. 41 The initial goal of these meetings was to discuss the ways in which women with lived experience of trauma could have their needs better met by primary healthcare. As the project developed, the group compiled some information about the project to encourage new members to join ( Figure 1). During initial conversations about the project, it was stressed that participation or the choice not to participate would have no bearing on the healthcare received and existing support from One25 would continue.
Meetings were attended by at least two professionals (GPs, researchers or staff from One25) to provide assistance, for example, if someone should become distressed and need to be supported within or outside the meeting. Over the period of operation (April 2019-to present) the core co-production group who met fortnightly has included a total of 29 women with lived experience of trauma, four researchers, two academic GPs, four GP trainees, and four One25 staff members, all of whom were women. Due to the longevity of the project, there has been significant variation in the membership of the group. At any one time there have typically been two to six women with lived experience, one to two academic GPs/GP trainees, one to two researchers and one staff member from One25 attending the groups. Group meetings happened on a fortnightly basis where possible, providing a sense of routine and continuity. Initial recruitment of women with lived experience of trauma MCGEOWN ET AL. | 1897 T A B L E 1 Comparison of principles of co-production and trauma-informed approaches.
Principles of trauma-informed approaches 9,21 Co-production principles 2,[18][19][20] Cultural, Historical and Gender issues Groups will seek to meaningfully engage with historic, gender and cultural issues in a way that avoids biases and stereotypes. They are aware of and respond to the needs of individuals from different racial, ethnic, religious and cultural backgrounds and different genders. Historical trauma is recognised and addressed, and the potential therapeutic value of cultural connections and gender-specific services is harnessed. 9 Accessibility, diversity and inclusion-Respecting and valuing the knowledge of all Covered by principles such as diversity and inclusion in social care practice literature. 19 Sometimes less explicitly addressed in some co-produced research literature, 3 but more in associated literatures on participatory action research and user-controlled research. 22,23 Different co-productive methods can be rooted in different critical theories, including Marxism, anarchism, critical race theories, feminism, disability rights and indigenous knowledge, 17,24,25 which highlight various structural inequalities and acknowledge how power is unequally distributed, creating boundaries and limits of co-production. 26,27 Not all co-production literature is rooted in these political approaches and there are more instrumental/tokenistic usages of the concept, which have been critiqued. 1,28 Peer support Peer support involves people with lived experience of trauma supporting each other directly. This can help some people process their trauma as they share difficult experiences in a supportive environment with other people who understand them. 9 Peer support and the co-production of services are integral to trauma-informed organisations, as relationships are based on mutuality and collaboration. 15 Building networks and peer support Reducing boundaries between professionals and people who use services. This may include more peer support roles, building networks and links between people who use services. Peer support action research is a co-production method that can be used by people with shared experiences to enable greater control over research design and knowledge generation. 29 Trustworthiness and transparency Joint understanding and clarity over roles Building and maintaining relationships Clarity of roles and how the process is led within co-production is important. 26 Building relationships is essential to co-production. 6 It is crucial to understand the temporality and dynamics of trust which means that trust needs to be continuously worked on through a collaborative approach. 30 Groups ensure transparency and consistency around how decisions are made, to maintain a sense of trust. 9

Collaboration
Partnership working is vital, with recognition that everyone working together has equal value and importance. Active efforts are made to identify and reduce power dynamics within groups and between staff and service users to demonstrate how relationships can be healing. 9 Reciprocity and mutuality. Embracing multiple perspectives, experiences and skills. Blurring roles and breaking down boundaries between professionals and people who use services 31,32 Using the knowledge of all those working together on the research, building on people's assets and the experiences they bring-everyone is of equal importance. Making sure the research team includes all those who can make a contribution, involving diverse stakeholders and being accessible and inclusive. 5,27,33 Empowerment and choice Sharing power Groups recognise the resilience and strength of those who have survived trauma. They seek to support people who have survived trauma to develop their own plans and goals, recognising their resilience and ability to heal and promote recovery from trauma. Understand how power differentials may diminish some people's voice and choice. Support shared decision-making. 9 The work is jointly owned and people work together to achieve a shared understanding, people are working together in more equal relationships. 2,26 There is a need to consistently reflect on power dynamics and make different forms of power explicit, reflecting on assumptions and usual ways of doing things. 26,34,35 Safety Safety Groups are in an environment that is not only physically safe from harm but also one in which all group members feel psychologically safe and supported. Interpersonal interactions encourage a sense of safety. Understand what is safe from people's own lived experiences and perspectives. 9 Safety is rarely listed as a core principle of co-production, although time to develop psychological safety with marginalised groups has been highlighted. 36 Participatory research has illustrated the need for safety for both academic and community researchers. 37 Other related issues include risk management and safeguarding. A risk-averse culture has been seen as a barrier to co-production. 38 Risk-averse cultures can be disempowering for peer-led co-production methods, 18 which can result in paternalistic professional dependency. 39 Risk awareness and a partnership approach where individual service users are involved in their own safety planning may be more conducive to co-production. 19 F I G U R E 1 Eligibility criteria and recruitment for Bridging Gaps. Design by Ellie Shipman.
took place in collaboration with One25, with L. P. and One25 staff directly approaching women who may have an interest in the project.
Some of the activities of Bridging Gaps over the 4 years of the project are outlined in Table 2.

| METHODS
Methods used for this article were based on participatory action research and cooperative inquiry, where research is developed with people rather than on them. 42 Women with lived experience were much more motivated by practice and the possibility of change, than by theory or research. Where they had been involved in research before, they often saw little change following this. The Bridging Gaps group were focussed on 'building participants' capacity to critique and question current arrangements, and innovate in the development of social practices', 17,p.86 aiming to improve access to traumainformed primary care.

| Ethics approval
The research was granted approval by the University of Bristol Faculty of Health Sciences Research Ethics Committee, references 93802 and 110882.

| Data collection
In traditional research, the roles of researchers and participants are mutually exclusive. Here all partners and women with lived experience contributed to the design and management of the work, and all were co-researchers and co-participants (e.g., researchers were also interviewed by more distant colleagues). 42 To understand, reflect on and document our approach to coproduction, we conducted the following data collection and analysis.

| Observation
Observations were conducted of three co-production group T A B L E 2 Bridging Gaps activities.
Received training from a communications and story-telling expert Co-developed and facilitated collaborative meetings with two general practices (face-to-face before the COVID-19 pandemic) Co-developed a training session on trauma for GP trainees (face-toface before the COVID-19 pandemic) Created a new process to enable public contributors to own the IT equipment they needed to be able to be involved. 41 (This was because most women we worked with were digitally excluded at the beginning of the pandemic). Received training in trauma-informed approaches Co-developed and delivered online training on trauma-informed approaches at two GP surgeries and three online codesign meetings (during COVID-19 pandemic). Co-delivered online training on co-production for the NIHR School for Primary Care Research. Supported four GP trainee placements within fortnightly meetings to improve access to trauma-informed primary care. Post-COVID-19 lockdowns, we facilitated three service improvement meetings at each of three general practices to improve access. Successfully implemented service improvements at two general practices, including the development of a bespoke clinic, with ongoing work with a third subject to funding (see Potter et al. 40   T A B L E 3 Illustrative data for cultural, historical and gender issues.

Themes arising Data
Cultural, historical and gender issues Quote 1: Well, what I like about it as well is because it's women who've been involved in-primarily, when it started, it was women who'd been involved in the sex trade, which I have, and I don't think them women's voices are heard enough; they're kind of forgotten about … the sex trade bit of it is kind of hidden and kind of looked down upon and I think it's really important to get them women's voices heard and make them feel that they are heard and have a voice in services because they don't have a voice in that enough. (Lived experience member 2) Quote 2: Women together and kind of just we have a bit of a giggle, have a cuppa and, you know, it just seems quite easy-going even though we're talking about quite-it can be quite traumatic, triggering things. Yeah, it feels like a safe space, definitely. (Lived experience member 1) Quote 3: I think women which have got experience at what they've gone through and the services that they are possibly using, to have their voices heard and to know that potentially you can make change is totally vital. Because for me, I was silenced for so many years. (Lived experience member 5) T A B L E 4 Illustrative data for Peer Support.

Themes arising Data
Supporting each other inspired by speaking to another member of the team who had talked to me about things she was struggling with … then I reflected on it myself and so it was the kind of bravery and strength that I see in women with lived experience who do share their stories that did prompt me to say 'okay … maybe actually I need to be brave and share some of my lived experience' and it was a real prompt for me seeking help for myself. (Academic GP 6) Impact of providing support to peers Quote 6: And then leaving and knowing they're going home alone as well, it stays on your mind. Well it stays on my mind for a bit … Yes, it's not nice to see that. Like one of the women broke down, yes and it was emotional. (Lived experience member 9) Quote 7: I felt like I had to stand up and be a kind of protector-not protector, that sounds like I'm making myself a martyr and I'm not-but I felt like I had to fight for them a bit more and fight for the kind of-to make sure things are all right for them a bit more. (Lived experience member 2) T A B L E 5 Illustrative data for trustworthiness and transparency.

Themes arising Data
Trusting relationships Quote 1: I think, because what helps with this, is a lot of the women know (Name), she's a doctor at One25 she does a clinic there, so a lot of the women would already know her, she would know the women and she understands the sex trade and what all that involves. (Lived experience member 2) Quote 2: The women have known her for a long, long time and she's done a lot to serve them really out of One25. That sort of trust, that's really hard to get in any other way because that's just through years of knowing people and showing up for them. (Academic GP 8) Quote 3: Just the level of trust that all women have really respected and all the women have really shared things as well that they wouldn't share so it's then made you feel like wow, like and if you all trust in me then I can trust you. If I'm trusting you, then you can trust me. (Lived experience member 9) Longevity of relationships Quote 4: Well. I think the longevity of it is really important … women felt that this was a long-term thing, that they weren't just involved in something for six months and then that ended. That was quite important in terms of women who have experienced a lot of trauma … It was quite important for there to be that time to build up that relationship and trust and then-because that takes ages and once that had been established you had time to deliver things through the project and women to feel, 'I am not just being used because of my lived experience.' (Support staff 7) Members contributing equally to the project

Challenges and disagreements on way forward
Quote 3: Creating a culture and a space where challenge can happen and conflict can happen and be resolved. That is to me the really interesting to learn about. (Support staff 7) Quote 4: We haven't had no disagreements as such. We have debates and I've walked off and I've gone away frustrated which I'm sure some of the GPs and that who are involved and (Name) have gone away frustrated from the way I've reacted sometimes but, as a group as a whole, we haven't had no massive disagreements or falling outs. We've managed to come to-once I get over the frustrationwe manage to come, we do come to an agreement and kind of work things out. Yeah. And I know everyone's thing is to make sure the women are safe and in a safe space, to be able to speak about things … (Lived experience member 2) 'Look, you don't understand. People will say "yeah" because they wanna please you and they're excited to be part of something and to have their voices heard but'-and it's not just on the staff, it's on the women as well and that's what I say to them, 'If it doesn't feel right, you're allowed to say no and no one's going to say, "Right, get out"', so I think it was from both sides as well, but if that understanding was there that that's-and they did, I explained it and we all got through it-but I think there wasn't that kind of real understanding of how service users can behave. (Lived experience member 2) Quote 9: I think you need to be working together for a really long time and there needs to be significant trust to say, 'I don't really like this', or 'let's go in a different direction.' (Academic GP 8) Quote 10: Yes a lot of people are like yes people and just like to agree 'cause they don't want to hurt your feelings or they don't wanna be challenged on anything or debated on anything. interviews, which were audio-recorded, transcribed, checked for accuracy and anonymised.  (Table 1) and a framework approach 43 was adopted in the subsequent analysis including interviews. These subsequent analyses were carried out using NVivo 12. Two researchers (T. S. and H. M.)

| Data analysis
double-coded two interviews, with subsequent discussion followed by T.
S. coding the remaining interviews.

| Bridging Gaps lived experience input
Bridging Gaps members have prioritised service improvements and achieving change in primary care over involvement in analysis and write-ups. In line with trauma-informed collaborative working, this focus was respected and supported while still offering opportunities to review drafts of key recommendations, learning and key points, which were discussed in our fortnightly co-production meetings. One

| RESULTS
Results are presented using the overarching framework of the six trauma-informed principles, with additional themes developed inductively and example quotes presented in associated Tables.  Over time, researchers also shared things about their personal lives with the group, which helped to engender a sense of equality and community in the group and avoid 'othering' (Table 4, Quote 4).

| Cultural, historical and gender issues
Listening to and working with each other also gave researchers a different perspective on their own help-seeking behaviour (Table 4, Quote 5).
We observed and discussed as a group how the peer support that existed between the women with lived experiences enabled them to challenge the researchers more confidently when there were things they did not agree with. Women would feed back that the peer support itself was healing, but listening and supporting each other also had an emotional impact: 'it was hard not to be able to take away the pain' (Lived experience member 9). We became aware that if group members took on too much of this peer support, it could become counterproductive for them and when this happened, we altered our ways of working to try and avoid this ( However, this added to her own concerns to make sure all voices were heard. On understanding this, the group changed its approach to focus less on trauma-informed training and experience-sharing, and more on developing bespoke service improvements with general practices.

| Trustworthiness and transparency
Ongoing relationships and trust with One25 staff from the start enabled a safe space to be established in which collaboration could take place. The relationship between the GP who worked at One25 (L. P.) who initially conceived of the project and women with lived experiences of trauma who had been supported by One25 was integral in promoting a sense of trust within the group (Table 5, Quotes 1 and 2). Trust was also built between women in the group through the sharing of experiences ( where some women who were in a place of greater instability attended the group, they could talk over others, or kept coming in and out through the meeting. This led to some disruption, which was difficult for those who were consistently and meaningfully participating (

| Collaboration
In the early stages of the project, there was more flux in group membership as individuals were more or less able to engage depending on what was going on in their life at the time ( Table 6, Quote 1). The lead academic GP reflected that this led to her holding group plans and newer members inheriting plans made by others who had left, which may have led to the perception that they were her plans as she was the one restating them.
Communication was essential to building collaborative relations.
Before the COVID-19 pandemic, all meetings were face-to-face and one funder required that the group attend training days to share learning with others. The car journey to these training days allowed an opportunity to deepen the relationships in the group. The neutral space of the car and no agenda for conversation may have reduced existing power dynamics, fostering openness and relationshipbuilding. In contrast, the COVID-19 pandemic had a significant impact on our ability to collaborate by preventing meetings in person.
We transferred to phone conferencing/online meetings and not all the women had access to e-mail/online conferencing initially, until researchers were able to provide IT equipment. When group members did have access to e-mail, people often did not want long MCGEOWN ET AL.
| 1905 emails as it could be too much information (

| Empowerment and choice
The project evolved from early phases where women spoke about their own personal experiences of trauma to a more focused problem-solving approach when collaborating with local general practices. Through initial group meetings, storytelling workshops and meetings with GP colleagues/trainees, we developed a series of activities to build confidence in interacting with professionals.  because there was something there that people could kind of access, but then that also has become quite difficult at times in terms of ensuring safeguarding and confidentiality, if other people are in the house and shared spaces, and not being able to guarantee and know who is in a space, which made it more difficult to kind of operate through lockdowns as we as progressed. (Researcher 4)

Differing perspectives on safety and safeguarding
Quote 7: On safety, we have to consider safety of people and we are responsible for everyone's safety and so there are just some things that we have to dictate. That's just how it has to be. (Support staff 7) Quote 8: When we were on Zoom there were some things around actually there might be a perpetrator there in the background … . I think … we were a bit concerned at the beginning about the team's understanding of safeguarding and actually the fact that women come across as really presentable but actually they have got all these things going on in their background and that that needed to be acknowledged. (Support staff 3) Quote 9: We were discussing about putting some safeguarding information onto an application form and it immediately evoked a response and she explained that response really clearly as to why even the terminology or any illusions to safeguarding was something that was quite traumatic and something that many women responded badly to who had had bad experiences under the banner of safeguarding. (Academic GP 6) External supervision Quote 10: I was a bit like, 'No, I think we really do need clinical supervision', and so then, fortunately, kind of linked in to get some … got an external person involved. So that was great, and I think that was really important. We've now funded, kept into the funding mix that there is that. (Researcher 4)

| DISCUSSION
While co-production processes are to some extent inherently trauma-informed, we identified various areas where additional considerations were necessary. The creation of a safe space is vital.
This should include direct consideration of cultural, historical and gender issues, which may impact upon group processes. The group should include people already known to and trusted by members who have an understanding of the signs and symptoms relating to trauma. Similarly, we found that when people agreed to ideas, 'yes' might not always mean yes, and on reflection, people might want to change their minds. An initial 'yes' could cover disagreement, uncertainty and fear, which rose to the surface at subsequent meetings. We encourage ourselves and others involved in co-production to explore how more safe room can be created on an ongoing basis for disagreement and uncertainty.
Partnership working is key due to the range of skills needed to carry out co-production work with people with lived experience of trauma. For those applying for funding, it is important that adequate resources should be provided for individuals to have flexible access to support between meetings from those with relevant expertise.
Additionally, funding applications should consider that time is needed for trust to develop, and maintenance of projects over months and years will be more productive than short-term projects.
Our recommendations (Table 9) are consistent with the experience of other co-produced research with people who have experienced multiple traumas, vulnerability or complex needs. 5,6,25,47 Moreover, we add to and highlight existing knowledge in the following ways: 1 That is valid, valuable and enough. They should feel supported to choose to illustrate a point with a personal example if they would like to, or not.
3. The challenging dynamics between safety and empowerment.
Safeguarding can sometimes mean power and choice is taken away from individuals, and this needs to be managed sensitively and carefully, and recognised and named transparently, explaining the issues and options for moving forward. Furthermore, this needs to be balanced with people's own expertise in how they manage their own risks. 48 4. Being honest about where things go wrong and discussing these openly.
5. Greater accounting for the needs of safe and sometimes singlegender spaces, or subgroups based on shared protected characteristics.
6. The importance of having a safe reflective space for all coproduction group members including people with lived experiences, support staff and researchers with external, independent clinical supervision available.

Reflecting on when you are blurring boundaries and why-what
impact is it having and is it helping to provide support and connection, or is it potentially adversely impacting on safety and trust?
Our theoretical contribution is that we explicitly compare and contrast co-production and trauma-informed principles, highlighting differences in safety, how boundaries are managed and maintained and how and when the sharing of experiences may be healing and empowering, or potentially retriggering. This point has significance for the 'deeper intellectual shift towards an epistemological position that values "knowledge as experience"', 49 and we highlight the caution and care that is needed to ensure that the sharing of experiences is healing and integrative, rather than potentially triggering of past trauma. Whilst sharing of experiences can help to 'work through emotional pain', 50 this process needs to be carefully supported and tailored to the individual at that particular time.
Our project highlights how researchers and practitioners working within these fields can use their social connections and power to help people with lived experience to access funding, resources and T A B L E 9 Recommendations table.

Trauma-informed principles and experiences/lessons learnt Recommendations
Cultural, historical and gender issues Having a women-only group enabled conversations to be held and experiences shared that could not have occurred in a mixed-gender group. Changing existing gender inequalities was a key motivator for individuals to become involved. We did not directly discuss racial inequalities within the group.
Additionally, the professionals within the group were from white backgrounds. This may have acted as a barrier to ongoing participation for women from a diverse range of backgrounds.
In advance of recruitment, reflect on the cultural, historical and gender issues that may affect your work. Work towards the co-production group being demographically representative of the population you are serving. Consider how cultural, historical and gender issues will affect power dynamics in the group and how to mitigate against this. Ask people in the group what they need to feel safe, people may prefer to be in groups based on gender or other identities/characteristics which could facilitate freer discussions and sharing of experiences. It may be helpful to discuss how cultural, historical and gender issues relate to the experiences of trauma and accessing care. Consider whether the recruitment of professionals on the team can seek to include representatives of a diverse range of backgrounds.
Peer support Support from peers had a therapeutic value for all the members of the group. At times, some individuals were providing a lot of support to peers which could have the potential to be overwhelming.
Build in time for supportive conversations within sessions throughout the project and acknowledge the potential therapeutic value of this. Where sharing of difficult or traumatic experience may arise, provide appropriate space for this. Consider the impact on individuals in the group who provide a lot of peer support to others and put in measures to proactively support them.

Trustworthiness and transparency
A prior relationship with those leading the project was vital in supporting engagement and enabling trust, particularly in the earlier phases. The long-term nature of the project enabled building of trust with new members of the group over time and maintaining trust of existing members that their efforts were making a tangible difference. Openness is vital to maintain trust. This includes the need for professionals in the group to admit and discuss when they get things wrong. We found funders to be very flexible where decisions came from people with lived experiences. Having agreed ground rules was important. Ensuring these were adhered to by all members of the group was at times challenging. Some individuals in the group contributed more than others for a variety of reasons. This did at times lead to a sense of unfairness from those who were participating a lot.
Seek to have key people who are known to and trusted by members with lived experience at all group meetings. Ideally, project funding should span years, not months. This is often difficult to achieve in practice within current structures. Lengthen projects to the maximum allowed by funders (i.e., span activity out rather than having short bursts of intense work). Be transparent about power differences and name them as issues arise, acknowledging that some cannot be overcome. Talk about things that have gone wrong and how to mitigate the impacts of these. Discuss decisions to be made and the parameters of these. Be transparent about how budget is spent and involve people in budget decisions. Explain to funders that where group decisions change project directions, that this has come from the group. Ensure ground rules are agreed upon by all members and shared at meetings. Identify who in the group is the most appropriate to ensure these rules are followed by all group members. Consider outlining a process whereby concerns can be raised about the behaviour of others in relation to the agreed ground rules. Consider as a group how best to manage differences in the level of contribution.

Collaboration
Collaboration between third-sector organisation, experts with lived experience, researchers and clinicians was necessary to bring all needed skills and expertise to the group. Academic language can be a significant barrier to collaboration on written reports and publications-consider creative ways in which to overcome these. Too many e-mails during lockdown led to some in the group feeling overwhelmed. Remote collaboration made it more difficult to maintain human connection and peer support. However, some in the group preferred remote access as they found attending in-person meetings difficult. Effective collaboration included different people expressing differing and conflicting views. Ensuring ongoing trust, transparency and a sense of safety within the group was vital to air different perspectives and find a new way forward to manage disagreements constructively. Some women with lived experience were able to contribute to the project at certain points but not at others due to various life factors.
Consider the range of expertise needed for the project and embrace a partnership approach. Ensure costing covers adequate reimbursement for the time and resources from all partners involved, including partner staff time and experts with lived experience. Use creativity to facilitate group contributions to written reports and papers-flip charts, screens, brain-storming sessions and post-its may all be helpful. Remote working and communication via e-mail may not be appropriate for everyone in the group. Ensure everyone feels safe to express how they wish to be communicated with and agree boundaries about this. Where possible, provide flexibility to allow different modes of engagement for different people. Ensure partnership working includes those with experience around communication and facilitation skills and experience in managing difficult discussions. Group facilitators should seek to ensure that all those in the group are heard and that the process for decision-making where differing views arise is fair. Where disagreements occur give significant time and space to understand these and be prepared to change direction as needed. (Continues)

Trauma-informed principles and experiences/lessons learnt Recommendations
Be flexible in your approach and provide opportunities for people to dip in and out of projects where they can. Some may wish to participate for a short time, others in the long-term. Offer a choice of in-person or remote participation where possible and give people the freedom to leave and rejoin the project based on changing circumstances.

Empowerment and choice
Engagement in storytelling workshops as a group had a team-building element, which effectively supported the development of trust within the group. Participation in co-production activities can be empowering for all those involved. Project involvement can support the development of planning, organisational, communication and research skills and specific expertise. Choices were provided about training courses to develop skills. We participated in the story-telling/communications workshops to develop communication and public speaking skills. At times, some women in the group did not feel able to say 'no' to plans that were being developed. We changed our approach to providing training so that people didn't need to share personal experiences.
Consider engagement in team-building exercises and social activities to strengthen relationships. This might include sharing experiences as a group if people wish to do this, but consider carefully how people can choose not to share, rather than feeling pressured to share. Think about how to support individuals in the group who are seeking to develop specific skill sets in the context of the project. For example, some may wish to play a more active role in producing research output, for example, papers, or presenting, while others might value playing a role in logistics or administration. Provide choices about potentially relevant training for people in the coproduction group and ensure this is costed into funding applications. Emphasise the freedom of individuals to disagree with the opinions of others, and to choose whether or not to share personal experiences. Provide opportunities for those in the group to feel safe to disagree with plans as they develop. Give opportunities for one-to-one discussions with trusted individuals to provide people the safety to say no. Ensure that it is made explicit from the start of the project that those with lived experience are under no obligation to share their difficult personal experiences, if they choose to share difficult personal experiences provide support and opportunities to pull out if needed.

Safety
Check-ins at the start and check-outs at the end allowed us to discuss our own emotional safety and signpost if support was needed. Not everyone has access to a safe confidential space at home, or IT or Wi-Fi, to access online meetings. Meeting online made it difficult to ensure confidentiality and provide support to each other. It can be difficult to predict what conversations might be triggering.
Some individuals might seem fine during the group but subsequently become more affected when on their own. These individuals were supported by the partner organisation. Differing approaches to safety and safeguarding need to be understood and strategies agreed upon between partners. Historical relationships/places may retrigger trauma and impact upon psychological safety.
Use check-ins and check-outs at all meetings to identify those in the group who may be struggling or need more support. Seek to provide a space for in-person meeting, where possible with flexibility for those who may wish to join remotely. Consider implications on confidentiality where meetings are being joined remotely and ways to maintain safety when meeting online. Ensure those leading the group have appropriate training to understand the impact of trauma, including symptoms of PTSD and complex PTSD, and strategies to support those who are experiencing these symptoms. Consider the need to cost in external clinical supervision for all those involved in the group and ensure the availability of support via partnership organisations between meetings. Ensure clear and transparent communication between group members around safety. Establish an agreed process at the start of the group about how safety concerns regarding group members should be managed and who should be involved in these situations. Preempt and explain who is likely to be at meetings so that people can discuss any issues arising and how to manage these. Be prepared for unexpected triggers and provide post-support wherever needed.

| STRENGTHS AND LIMITATIONS
Our study's strengths include that it uses both a data-led and reflective practice approach to identify recommendations and adaptations for those developing trauma-informed services using co-production. Despite the growing interest in trauma-informed services, there has been little attention given to date on the processes of partnership working between professionals, researchers and people with lived trauma experience as they develop services. Without careful consideration of these processes, there is a danger of partnership working being tokenistic at best, and even perpetuating existing power dynamics. 28 This is unlikely to lead to the transformational systems changes necessary to tackle existing inequalities. Future work could include further co-produced service improvement projects based on our recommendations, with further evaluation.
Our project took place over a number of years, which allowed time for reflections and lessons to be learnt. Interviews included a diverse range of perspectives including those from clinical, research and voluntary sector backgrounds and those with lived experience of trauma. Analyses were carried out jointly between those who were involved in group processes and those who were external, enabling triangulation between different viewpoints when analysing the data. Observational and reflective data were triangulated with interview data. Although the pandemic led to many difficulties in our group processes, it provided some valuable learning on the opportunities and difficulties of remote working for those with lived experience of trauma.
One limitation was the fact that changes in group membership meant that many women who were involved for certain periods of the project could not be interviewed about their experience. While those interviewed suggested that these women dropped out for reasons unrelated to the project, it would have been helpful to have their perspectives and reflect on whether any additional support may have enabled them to remain part of the process.

| Looking forward
At present, the future has challenges and opportunities ahead for Bridging Gaps. Opportunities include some further Bridging Gaps funding to develop a website with our resources for general practices and to reach out to more general practices. Academic GP (L. P.) has received doctoral funding to develop a complex intervention to improve access to general practice for people with severe and multiple disadvantages. 54 Through these activities we are linking with others who have similar missions, to explore potential collaborations.
Challenges are that our partner charity One25 has experienced significant financial difficulties and has had to reduce its provision, closing its drop-in for women who are street sex working or at risk of street sex work, and support services for women who are further on in their recovery. 55 Ongoing challenges for securing funding in the context of a cost-ofliving crisis where support services are in even higher demand is a situation that more and more charities are facing. [56][57][58] These issues highlight the ongoing challenges that can be faced with short-term, project-based funding when needs can be complex and long-term. The implications of this cut in service provision have been of great concern to Bridging Gaps. All lived experience members have received support and encouragement from this organisation and came to the Bridging Gaps project by way of it.
That such opportunities will no longer be available to vulnerable women, as they were in the past, has been deeply frustrating and upsetting, not just for women with lived experience but also for the professionals as they will be dealing with the reality of these closures. Bridging Gaps has become more than just a 'research project' for members, it is a trusted place of safety, of nonjudgemental support and togetherness. However, this will be harder to manage in future as needs become greater because of the closure of their support network.

| CONCLUSION
Our findings provide vital learning points for all those seeking to develop trauma-informed services and an opportunity for further evaluation of our recommendations for practice. The high prevalence of trauma in the general population also makes our findings even more broadly applicable and merits consideration for all those engaging in co-production work.

CONFLICT OF INTEREST STATEMENT
Maria Carvalho and Florrie Connell worked for One25 at the time of this project. One25 received a fee to compensate the charity for the time that their staff committed to the project.

DATA AVAILABILITY STATEMENT
Original data are not available and research data are not shared due to the potential to compromise people's anonymity and the small number of interviews that were undertaken.