‘No‐one has listened to anything I’ve got to say before’: Co‐design with people who are sleeping rough

Abstract Background Despite policies and programmes aimed at housing people who are homeless, there are still people who live and sleep rough. This project used the skills and knowledge of people in this situation to identify a strategy to mitigate some of the risks. Objective To describe the development and conduct of a co‐design project involving people who are homeless. Setting/Group Members A Working Group of 11 was formed following a careful recruitment process from people who had volunteered after consultation by the project team. The co‐design approach was guided by a set of principles. Methods Eight members of the Working Group were interviewed by an external researcher (RM). The approach was primarily deductive, with the principles adopted by the project team used as a framework for data collection and analysis. The co‐design process was captured by the project leaders (BK, PC) supplemented with documentation review and team discussions. Results The group met weekly for 12 weeks, with 8‐10 members present on average. They reviewed information from the survey, contributed ideas for solutions and ultimately decided to provide information via print, a website and an event. Important factors in on‐going involvement were carefully selecting group members and making participation rewarding for them. Discussion/Conclusions Vulnerable people such as those experiencing homelessness can be excluded from decision‐making processes affecting them, as they can be perceived as hard to reach and unable to make a meaningful contribution. This project demonstrated that a carefully managed project, with sufficient resources and commitment, it was possible to involve people who are homeless and maintain involvement over an extended time period. Public Contribution The Working Group reviewed survey findings and developed an intervention to minimize the health, social and legal harms of sleeping rough. Several members reviewe this paper.

lived experience of homelessness to develop strategies that may improve health and well-being is required because of the specific nature of the knowledge and the negative consequences of lack of information.
Even in high-income countries such as Australia, people who are homeless have been shown to have high rates of morbidity and mortality, including substance use issues, communicable and noncommunicable diseases, and psychiatric problems. 12,13 They also tend to seek care late, once health issues become emergencies, in part because the efforts involved in meeting basic needs such as food, water and shelter take priority. 14 Despite recognition of the barriers faced by this group, we could not find any examples of codesign undertaken over a period of months with people who are sleeping rough in the refereed literature, as opposed to one-off consultations or co-design not directly involving the homeless. 15,16 The goal of this paper was to evaluate a project that used codesign principles to utilize the skills, knowledge and experiences of people with a current or recent experience of rough sleeping to identify some of the health, social and legal issues faced when sleeping rough and strategies to address them.

| OVERVIE W OF THE PROJEC T AND RE S E ARCH
The project was initiated by cohealth which is one of Australia's largest community health services, which has a mission to improve health and well-being for all, and to tackle inequality and health inequity in partnership with people and their communities.
A two-person team was appointed to manage the project, a Project Worker (BK) with a social work background and a Peer Worker (PC) who has a lived experience of homelessness, is a graduate of the Council to Homeless Persons (CHP) Peer Education Support Program and who was employed at cohealth in an Allied Health Assistant role. BK and PC are also co-authors on this paper.
The background work prior to the co-design project is not reported in detail in this paper, but in brief, it comprised developing and collecting information using a survey which was completed by 81 people with current or recent experience of sleeping rough. 17 Cohealth's internal Human Ethics Advisory Group approved the project (HEAG 1902), consistent with the organization's policies for internally initiated data collection without the intent to publish.
After the consultation data were collected, a Working Group was put together to review the survey findings, identify the key issues faced by people sleeping rough and consider strategies to support people to stay healthy and safe. The group met weekly for 12 weeks from September to December 2019, facilitated by BK and PC.
PowerPoint slides were used to structure the sessions. The format of each working group meeting was: Australia, co-design, homeless persons, housing, qualitative, social problems • Acknowledgement of Country (a statement that shows awareness of and respect for the Traditional Custodians of the land).
• Reminder about the group agreement (see below).
• Start-off activity such as poems, mindfulness exercises.
• Updates including things such as other opportunities for involvement, current campaigns and things happening within homelessness, as well as information about things the facilitators had been doing between the workshops.
• Reviewing the findings from the survey, with discussion and reflection based on presentations made using graphs, quotes and thematically analysed data.
• Activities including small group work and large group work, for example matching aims to proposed ideas, matching findings from the survey to ideas about how to respond, developing an idea prioritization matrix, narrative activities and storyboarding.
• Closing comments, including 'what's on next time' and at key timepoints, evaluation questions.
Within this overall structure, each session had a specific focus: 12. Agreement on producing something to share information peer-to-peer, and storyboarding how it would be put together Some core guiding co-design principles were adopted by the programme staff when developing this project. These were based on literature 18,19 particularly principles from the Co-design Initiative, as well as personal experience: • Inclusion: recruiting a diverse group of people with a recent lived experience of rough sleeping.
• Equity (and respect): ensuring that all voices in the group were heard and valued.
• Capacity building: supporting people to participate safely, and in a way that is empowering.
• Purposeful: working towards a useful outcome.
Other key principles from the Co-design Initiative of 'co-created', 'innovative' and 'evaluated' relate to the product which resulted from the co-design process, which is out of scope for this paper.
The final outcome was a zine (an informal magazine, selfpublished and frequently produced on a photocopier) which was made available in print and online (http://www.needt oknow homel ess.org.au/) as well as a website and an event where information could be read out. It contained information, stories, encouragement and advice that could help people who were sleeping rough, for example advice for staying healthy on the streets. Subject to funding, the plan is to produce it regularly.
During the co-design process, independent university researchers were asked to evaluate some aspects of the project, specifically the way the Working Group had been put together and run. They 3. The views of the facilitators about the co-design process.

| Interviews with the group members
In June 2020, all members of the Working Group were contacted and asked to take part in an interview. Three of the original members were no longer actively participating and all of these declined to be interviewed. Eight interviews were conducted: five face-to-face and three by telephone. The face-to-face interviews were conducted at cohealth, next door to where the Working Group meetings had been held, to ensure the interviewees were as comfortable as possible.
Written consent was obtained at face-to-face interviews, and verbal consent was recorded for telephone interviews.
The four co-design principles used to develop the project formed a framework for the semi-structured interview schedule. The questions focussed on the group members' motivation to be part of the Working Group, their experiences of being selected and then being part of the group, things they felt could be changed about the overall process, and their satisfaction with the zine (Appendix A).
Interviews were audio-recorded and then played back while detailed notes were taken. The analysis approach was primarily deductive, guided by the co-design principles that provided the framework for the project's development. The analysis method followed the general approach suggested by Braun and Clark 20 with notes read over frequently for familiarization and initial identification of patterns (RM). Themes were reviewed and confirmed through discussions between RM and VL. BK and PC were not involved in the analysis of the interviews, as they would have easily been able to link the interview to a specific group member. Group members were assured their interviews were confidential, and all quotations use gender neutral pseudonyms, to help protect their privacy. Quotations included in this paper use ellipses to remove irrelevant asides and unnecessary words, 21 but are otherwise verbatim.

| Other input
Documentation collected throughout the project was reviewed, including the PowerPoint presentation used in each session.
Additionally, the facilitators have contributed their perspectives and are authors on this paper (BK, PC).

| Recruitment to the group
Initial interest in being part of the Working Group was gauged from those who completed a survey as part of the project's initial consultation of people experiencing homelessness. At the end of the survey, respondents were asked if they were interested in participating in a working group which would 'discuss the responses to the survey and design and develop something that aims to reduce the health, legal and social impacts of sleeping rough'. They were advised that this would involve committing to 12 weekly 2-hour group sessions for which they would be paid. Sixty-one of 81 people surveyed expressed interest at the time they were surveyed. A multi-step process was then undertaken to select members of the Working Group.
1. Contact was made with those who had expressed interest in participation, either by telephone, email or via information to workers or services to pass on. At this point, people were given more information about the project (eg the weekly commitment needed).

2.
A more formal telephone interview was conducted with potential candidates who were paid for their time. This covered their motivation to be part of the group, skills and experiences they would bring, their ability to get on with others and to self-regulate their emotions. Availability to attend and support needed to do so (eg public transport cards) were also assessed.
3. The information from the interviews was discussed by BK and PC to develop a short-list of potential members. This took into consideration socio-demographic factors such as gender, age, ethnicity, LGBTQI and Indigenous backgrounds, length of time homeless/rough sleeping as well as the interest and capacity to participate identified in the telephone interviews. They were also asked how comfortable they would be discussing their own lived experience in the group.
4. The key candidates were interviewed face-to-face and final decisions were made based on consideration of the mix of personalities, how well they would work together, and how they could be expected to handle the experience of discussing potentially distressing content.
The initial group included seven men, four women and one nonbinary person; four were born overseas with two culturally and linguistically diverse; four identified as LGBTQI and one as Indigenous.
They ranged in age and rough sleeping/homelessness experiences.
One demographic which was not successfully recruited, despite extensive efforts, was young people.

| Inclusion: Joining and staying in the group
The Working Group members described their motivations for joining the Working Group primarily as the desire to share their insights and to help others who found themselves on the streets. A couple described feeling overwhelmed and happy when it was confirmed they would be on the Working Group. Although altruism was key, the financial incentives and other support offered also contributed to motivation. Some group members were clear that they would have participated even if they were not reimbursed for their time, though others felt it was appropriate that they were paid for their time and their skills and said that it also increased their sense of what they were doing as being their 'work'. The provision of lunch and transport cards was welcomed, but more as part of feeling respected and valued than a key reason for being here.
Attendance at each group usually ranged from 8 to 10 people.
At the time of the interviews, approximately 9 months after the first meeting, eight people remained actively involved. The eight who were interviewed had maintained their enthusiasm or become more motivated.
[I] probably got more motivated, because I was excited to see friends that I'd made, have that chat, have that discussion…towards the end it was more exciting to go knowing the environment I was going to be in.

Jamie
It's not important, but what I liked was the welcoming when you got there, when everyone got there.
Everyone was nice, like hello to everyone. For me I don't usually get that… Everyone was genuine, not being nice cos they wanted to get credit off you.

Charlie
The facilitators identified some practical aspects of the meetings which made them enjoyable for the group members: tea, coffee, a nutritious lunch and snacks were supplied, as well as a selection of fidget toys (stress balls, Rubik's cubes) and weighted blankets and cushions which group members could access at any time. One group member remarked to a facilitator that it was a highlight just to be in a warm room with food.

| Equity and respect: ensuring that all voices in the group were heard and valued
There was a strong sense of being respected and made comfortable by the facilitators and other staff at the community health service where they met.
Way we were treated…as equals. I think you expect to walk into a place like this, they're professionals, this is a professional place, we've come in off the street.

Sam
Being heard and valued did not relate directly to being agreed with; not all group members believed that a print resource was the outcome they wanted from the group. Some were disappointed this was of more benefit to those who were new to the streets rather than the current homeless who 'already know this stuff' (Ash), but others saw this as a valid thing to do 'Knowing there was very little information available to the homeless on the streets, it was way for me to give back, while I still had a clear mind' (Jamie). But regardless of their feelings about the end-product, group members unanimously said they believed they had been heard.

| Capacity building: supporting people to participate safely, and in a way that is empowering
An initial task for the group members was to develop a Group

Chris
We did really well with it. The main thing was lots of us like to jump in…Needed to learn to let people fully speak, get their opinion out before we spoke. We all did really well, stuck to the agreement.

Ash
Some group members were able to list very specific skills they had developed directly as a result of being part of the group, such as public speaking, or opportunities to be on other advisory committees. For most, the benefits they described were more around self-confidence, social connectedness and making friends in an enjoyable setting.
A key thing the group members associated with being in the group was confidence, whether it was gaining or regaining confidence.

| Purposeful: working towards a useful outcome
In the first session, the purpose of a Working Group was made very clear, so that members understood it would be time-limited and with a goal to achieve-in this case devising something which would help people to live more safely on the streets in the short term and help them to eventually link to services and supports over time that could help them avoid entrenched homelessness. The group took part in extensive activities such as reviewing the data from the surveys, developing Idea Matrices and discussions about realistic time frames and costs for an end-product.
Some of the group members recalled that the group had always been set up to produce a printed resource of some kind: 'a useful true fact pamphlet with 100% facts in that people need' (Charlie) but others recalled the initial brief was much broader: We were the voice of all homeless people, not just ourselves… to make something from a lived experience perspective rather than a worker perspective… wasn't until about halfway through we came up with the idea of the helping out booklet.

Alex
The decision to produce a zine/website/event disappointed some group members who would have preferred their ideas be taken up, which primarily revolved around strategies to improve access to storage and personal safety.

| Discussion
This project demonstrated that it was possible to involve some of society's most vulnerable people in an extended co-design project which was sensitively managed and appropriately resourced. This was a far more intense process than one-off consultation or focus groups-which in some cases can be tokenistic-and given the dif-

| Essential elements in success
There were three key elements to successful application of the codesign principles used in this project.

| Selecting appropriate group members
Inclusion is a core principle for successful co-design; however, it can be difficult to attract marginalized groups to become involved in public participation groups. 8 One suggestion is to recruit through established networks 7 ; however, in this project, initial interest in participating in the Working Group was gauged from those who completed a survey, thus spreading the net well beyond known networks. The high level of interest then allowed the facilitators to use a detailed, multi-step process to select group members.
The selection process went beyond choosing those with particular experiences or from specific demographics, or who had been involved in similar groups. The group was not restricted to those who had limited experiences living on the street or were generally coping well, they had extensive and varied histories of rough sleeping. The selection process allowed for people to be selected for their willingness and ability to commit to and contribute to the group. A diverse group of participants joined and stayed in the group.

| Making participation a positive experience
After recruiting the group, it was important to conduct it in a way which maintained interest, enthusiasm and on-going participation.
Every stage of this project was designed to ensure that the group members benefited in some way. Those who took part in the survey were reimbursed for their time, as well as being given other tangible items such as meal vouchers. More importantly, they were connected with services if it was appropriate. This kind of support goes beyond capacity building; it reflects the principles of the community health organization, which offers holistic care including referrals for diverse needs as they emerge.
The prime motivation described for being on the group was the desire to share insights and to help others who found themselves on the streets. A couple of group members described feeling overwhelmed and happy when it was confirmed they would be on the Working Group. Other studies have also described altruism as a motivator, but not with such a disadvantaged group. 22 This group expressed a strong sense that their voices were being heard and valued, and for many, this level of recognition was a novel experience, reflecting the presence of the principle of equity and respect.
A crucial component in supporting safe participation was the agreement that was drawn up by the members, which clearly spelled out what was expected of them in their interactions within the group. They took total ownership of the agreement and saw it as a way to help manage both their own and others' behaviour while in the group. Within each session, the group agreement was presented as a PowerPoint slide, and often read out, which served to remind people of how they had agreed to behave. Importantly, slides were also used to regularly restate the overall goal of the project as well as goals for specific meetings. These factors helped with setting boundaries and keeping the project on track. Some group members were keenly aware of their own difficulties and lack of experience in working in groups-as well as that of others'-and were grateful for the reminder about how to behave.
The less tangible aspects of being in the group were also crucial to the co-design process. The members developed a strong sense of camaraderie as well as connections with the facilitators. They described feeling welcomed when they arrived, they chatted informally and shared information among themselves. The group was also run in a way which enabled members to withdraw if they were feeling upset or overwhelmed. They could have tea or coffee at any time or get a stress toy-which could also act as a signal to the facilitators that they may need some help. The facilitators used their experience as health professionals to monitor the health and well-being of group members. They would debrief after each session and identify any concerns that could be addressed by referral or assertive outreach.
Where there were concerns about group members' well-being, the project workers would follow-up by phone between meetings. In addition, the meetings were deliberately conducted next door to a community health service, to encourage engagement and use the of health services available.
Several group members commented on the on-going benefits of their experience in the project on their lives, reflecting achievement of the principle of capacity building that is empowering. The facilitators noted an increase in empathy, willingness to listen to others, and a more positive outlook by group members as the group progressed.
One facilitator observed that planning tasks, keeping to a timeline,

| Outcomes of the co-design process
The concrete outcome of this group was a three-pronged informa- Rather than choosing people who were more stable, such as only those who were recently housed, the choice was made to include people with current or recent experience of sleeping rough because of the relevance of their knowledge and experience. The facilitators recruited a few more people than they thought were needed to allow for this natural attrition, as they did not want to damage the group dynamic by recruiting additional members at a later stage.

| Power differentials
Co-design projects raise ethical issues 24 and recruitment, maintaining engagement and managing power differentials are particularly challenging when projects involve vulnerable people. 8  rather than improving an existing one; therefore, they were not in the position of directly criticizing cohealth.

| COVID-19
The advent of COVID-19 meant that distribution plans for the zine needed to be adapted, as some services shut down, and many rough sleepers were moved temporarily into hotels or other accommodation. The group determined alternative strategies for distribution such as including copies in essential item packages distributed to people on the streets and outreach teams from the community health organization to taking the zines to hotels. The content was also rapidly modified to include a supplement about COVID-19, demonstrating the value of having on-going mechanisms for working with vulnerable populations in crises, as recently suggested. 25 Meetings of the group were also impossible during a hard lockdown in Melbourne, and issues around the digital divide are very prominent in this group. 26 Although most people who are homeless have basic mobile phones, few will have access to software or skills to participate in online meeting platforms. Thus, contact became predominantly one-on-one between a facilitator and group member.

| CON CLUS IONS
This project demonstrated that very vulnerable and disadvantaged people were willing to take part in a co-design process that involved multiple group meetings over more than 3 months, and that they welcomed the opportunity to have their voices heard in a safe setting and their ideas put into action. The selection of the group members and the creation of an environment where they felt valued and nurtured was crucial to allowing them to share their voices and actively create practical resources. As one person summed up: 'No-one has really listened to my side before, anything I've got to say before. Now I've done this'.

ACK N OWLED G EM ENTS
The authors would like to thank the members of the Working Group, particularly those who took part in the interviews: Andrew Rushworth, Pedro, Yasmine Jensen-Solyom, Ken Jones, Alicia F, Andrew McVeigh, Brett Berry and Jeffrey Steet.

CO N FLI C T O F I NTE R E S T
The authors declare no conflict of interest.