Recovery in Mind: A Recovery College's journey through the Covid‐19 pandemic

Abstract Introduction The Covid‐19 restrictions of 2020–2021 are known to have undermined the UK population's mental health. Working alongside staff, peer trainers and students at Recovery in Mind (RiM), a Recovery College (RC) in West Berkshire, England, this mixed‐methods study is amongst the first to investigate how an RC has responded to the pandemic. Methods Working in co‐production with RiM staff and peer‐trainers, this study employed a mixed‐methods design, gathering Warwick‐Edinburgh Mental Wellbeing Scale (WEMWBS) well‐being outcome measures by questionnaire and student experience, learning and co‐production by interviews. Findings This research found that RiM continued to produce demonstrable improvements in student mental health. Students welcomed the way that RiM adapted to offering online and socially distanced provisions. Students valued the skills that RiM taught and the way that RiM courses reinforced prior learning; above this, however, they valued the mutual support and sense of community that participation provided. Conclusion This study underlines the value of RCs maintaining ‘hidden curriculums’ of peer support and community involvement. This research emphasizes co‐production as not only a tool for empowerment or service improvement but as a valuable skill for personal mental health recovery. Even when operating under the most unforeseen or challenging of conditions, RCs should always endeavour to prioritize and maintain co‐production. Patient or Public Contribution In accordance with the RC ethos, this was an entirely co‐produced study, with academic researchers and RiM staff and peer trainers working democratically in partnership with one another to design and manage the study and to write up and disseminate findings. To ensure the independence and rigour of findings, data analysis was undertaken by external academic researchers.


| INTRODUCTION
The Covid-19 pandemic and its 'social distancing' restrictions are known to have undermined population mental health, both globally and in the United Kingdom. [1][2][3][4] During 2020 and 2021, the entire United Kingdom was placed under national lockdown three times (March-June 2020, November 2020 and January-March 2021), with a complex tiered system of localized restrictions between these. The pandemic has led to increasing symptoms and declining well-being for people with pre-existing mental ill-health. 5,6 Both people with and without pre-existing mental health needs reported finding mental health support more difficult to access and engage with. 5 To date, little research has evaluated how Recovery Colleges (RCs) or similarly user-led or co-produced mental health services have responded and adapted to the Covid-19 pandemic. 7 However, emerging evidence both globally and from the United Kingdom suggests that, from the outset of the pandemic, co-production, broader forms of service user participation and patient and public involvement 8 in mental health services have been significantly deprioritized over the last 2 years.

| RCs
In the United Kingdom, the RC model was developed in the mid-2000s. 9 As of 2020, there were approximately 80 RCs across the United Kingdom, with an increasing number operating worldwide.
The RC model is guided by values of hope, control and opportunity, and aims to empower individuals and communities with the skills to improve their own personal and collective mental wellbeing. RCs seek to function as educational rather than therapeutic institutions: participants are students not patients; students register for courses rather than being referred, and students attend classrooms rather than day centres. 10,11 RCs adhere to an educational philosophy of emancipation and action learning, 11,12 grounded in co-production by valuing the lived experiences of people living with mental health challenges. 11,13 RC co-production involves students and peer trainers. 11,12,14 Peer trainers are former RC students who have been employed to coproduce all aspects of RC planning, development, governance and delivery, drawing upon both their lived experiences of mental illhealth and recovery, 11,14 and upon their broader personal and vocational skills. 11,15 In the same way that the model differentiates RCs from organizations providing treatment or care, so also it differentiates peer trainers from many of the peer support workers increasingly working within other forms of mental health services provision. Peer trainers are educators and mentors, not carers, and may exercise their skills at all strategic and operational levels of RC activity. 11,14 The co-production of RCs, therefore, empowers people with lived experience not only to contribute and become involved but to assert leadership and enact change. 11,13 RCs have been extensively evaluated and researched, 10,[15][16][17][18][19][20][21][22] with systematic and narrative reviews offering a strong evidence base from which to claim that RCs are effective in improving service users' quality of life and self-management skills, and lead to lasting organisational change within mental health services. 10,23-25 A Canadian initiative is currently co-producing a user-led framework for evaluating RCs. 26 Increasingly, RC research is diversifying, exploring the experiences of students with specific diagnoses and studying the impacts of specific RC courses. 21,27,28 One of the first pieces of RC research to have explored the effects of the Covid-19 pandemic and the move online is that of Rapisarda et al. 29 Their study found online RCs to be particularly effective at combatting anxiety, and advocates RCs as a cost-effective and accessible public mental health response to crisis situations. 29

| Recovery in Mind (RiM)
RiM is an RC in West Berkshire, in rural South-East England. It was founded in 2016 and is led by AR, an experienced voluntary sector manager with lived experience of mental ill-health. In addition to AR, RiM's team comprises Occupational Therapy (OT) staff and peer trainers. Some of these OTs are seconded by the local NHS Trust and some are employed on a zero-hours basis funded by Big Lottery Fund. Peer trainers are also given zero-hours contracts which, like the salary of AR as the RiM CEO, are funded by various UK charity grants.
Everyone on a zero-hours basis is paid in excess of the local living wage and has been fully remunerated for work they would have done if not for the disruption of Covid-19.
The study aimed to evaluate RiM's impact and its courses in promoting students' well-being and subjective perception of recovery and to explore students' and peer trainers' lived experiences of recovery through participation in RiM. The study was integral to the RiM's expanded course delivery supported by Big Lottery Fund, and lasted over a 2-year period, from June 2019 to June 2021. This study also served to document RiM's journey through the pandemic. This paper explores how and why RiM has YOELI ET AL. | 3275 adapted its courses over the pandemic and how students and peer trainers journeyed with RiM.

| Charting RiM's journey through Covid-19
More than 110 students joined RiM Steps 1 and 2 courses over the study period of June 2019 to June 2021.
In March 2020, with the first lockdown in the United Kingdom enforced, RiM's initial response was to start a weekly email newsletter to communicate with all current and former students and peer trainers.
With the lockdown and restrictions on face-to-face contacts continuing to render it impossible to run courses for months, RiM converted the Step 1 Recovery Bitesize into a video. RiM staff produced a series of online nano-courses and relaxation exercises as a freely accessible resource to all struggling to cope with the pandemic. Essentially, the team tried to remain flexible in their approach, and this, coupled with a 'can do' attitude enabled the service to respond to the pandemic. RiM came to regard the use of technology as essential in keeping communication going despite the initial reticence and anxiety from staff, peer trainers and students alike. The team had to make a number of 'best guess' judgments in the early days of the pandemic when online communications were still new to many. In January 2021, the team launched a regular online Coffee Catch-Up session.
With the partial lifting of restrictions in the Summer of 2020, RiM experimented with outdoor activities built upon existing nature-based provisions. Particularly successful were the Walk and Talk sessions, later renamed Ponder and Wander. In Autumn 2020, RiM resumed indoor face-to-face interaction by delivering Welcome to Recovery (WTR) (see Figure 1) in a larger and more Covid-secure venue than previous premises. When the November 2020 lockdown was announced shortly after the start of a Five Ways to Wellbeing course at another outdoor venue, the RiM team decided to move the remainder of the course online using Microsoft Teams. In December 2020, RiM resumed some outdoor sessions, culminating in a physically distanced Festive Walk.
Throughout the pandemic, all face-to-face activities were However, when Recovery Bitesize workshops were delivered as an emailed video link, no questionnaires relating to it were administered as it felt inappropriate without a face-to-face opportunity for the team to explain the research.
The three-stage model of RiM 3.2 | Co-production of the mixed-methods approach When studying user-led or co-production initiatives, it is necessary for research to appropriate and extend organisational valuesin the case of RCs, those of hope, control and opportunityto amplify the voices of the historically least-heard participants, and to challenge their marginalization. 30 This study was co-designed and co-produced by researchers, RiM staff and peer trainers, who developed a mixedmethods approach aiming to combine the advantage of quantitative and qualitative research (see Figure 2). 18 34 2. A two-part Personal Goal Questionnaire was devised by RiM staff and peer trainers to establish a subjective evaluation of the achievement of a personal goal. Each student specified a personal goal that they would like to achieve through the RiM courses, followed by a Likert scale of 0 (worst) to 10 (best) to rate themselves in terms of achieving that.
The researchers invited each student to complete WEMWBS and a Personal Goal Questionnaire at four time points: T0 (before their Step 1 Recovery Bitesize workshop); T1 (at the beginning of their Step 2 Welcome to Recovery (WTR) course); T2 (at the end of their WTR course) and T3 (3 months after they had completed WTR). When courses were delivered in person, questionnaires were completed by students at the beginning (T0 or T1) or the end (T2) of the respective session. T3 questionnaires were posted to students with a selfaddressed stamped envelope to return completed assessments.
However, as previously explained, when Recovery Bitesize workshops were delivered online, no T0 questionnaires were administered.
During the time of online delivery, questionnaire administration was via post and email, with telephone support provided for those who preferred to complete questionnaires over the phone.
In addition to comparing students' mental well-being with the English population norms on the WEMWBS 34 using independentsample t-tests, we explored changes in students' well-being and goalachievement across the time points, using descriptive statistics.
Subjective ratings of personal goal achievement were analysed descriptively. Missing data on individual WEMWBS scores were F I G U R E 2 Flow chart of study design YOELI ET AL. | 3277 dealt with by replacing up to three missing and incomplete WEMWBS items with the average value of the other item values. 35

Sampling and recruitment
When the research was first designed in 2019, researchers envisaged a purposive approach within which the researchers and RiM would work together to select and recruit a sample of students who might best reflect the diversity of student demographics. However, as Covid-19 disrupted course delivery and study recruitment, researchers moved towards a more pragmatic sampling method, inviting participation from students who had remained most engaged with RiM following its move to online provision.

Interviews
Semistructured topic guides for the student interviews were designed and piloted by RiM staff, peer trainers and students, working together in accordance with the co-production ethos of RiM.
Student interviews were carried out by telephone and Zoom by SB, asking participants about their RiM experiences and more generally their lived experiences of the Covid-19 pandemic and recovery.
Interviews were audio-recorded, transcribed and subjected to thematic framework analysis 36 using NVivo12 research software.

| Ethics
Ethical approval for this study was obtained from the University of

| Subjective evaluation of personal goal achievement
At the beginning of the courses, 73 students completed the personal goal achievement question. Some of the wide-ranging personal goals the students set for themselves are detailed in Box 1.
The baseline goal-achievement scores ranged from 0 (not at all) to 10 (completely), with a mean score of 3.07 (SD 2.16) and a median score of 3.00. These scores indicated the students felt they had a long way to go in achieving their goals. At T2, 38 students rated their own personal goal achievement. For male students (n = 8), the mean score was 5.63 (SD 2.07) and the median score was 5.50. For female students (n = 30), the mean score was 5.03 (SD 2.08) and the median was 5.07. At T3, 16 students, 5 males and 11 females completed this question. The mean and median scores among the male students were 4.20 (SD 2.49) and 4.00 and for females were 5.18 (SD 2.71) and 6.00, respectively.

| Interview findings
Six students were interviewed, in interviews ranging from approximately 25 to 90 min duration. The analysis identified three main themes, each with a number of subthemes.

| The skills that RiM taught
In describing the contents of their personal toolkits, students distinguished between the skills that they had been taught within RiM classes and the wellbeing strategies which RiM had empowered them to discover and develop for themselves. The peer support helped us to develop and learn

Being supported during lockdown
Particularly during Covid-19, students valued the opportunity they provided for conversation, and the connectedness they offered to those who could not attend the socially distanced outdoor sessions: [online] coffee mornings, they were good. Coffee

BOX 2 A student's journey: Heidi
Heidi works with racehorses, but is currently off work following a series of work-related injuries. One of her physiotherapists once had a similar injury to hers, and it was hearing about his experience that provided her with an introduction to the benefits of peer support. Heidi has struggled with her mental health for some time, and has had several therapists, one of whom, due to Covid, she has seen entirely on Zoom. Whereas she has always found therapy helpful, she has found that RiM has taught her a lot of things that therapy has not, particular in terms of the practical skills for well-being.
So far, Heidi has been the youngest student in both of the online courses she has attended. This does not matter to her at all-instead, she has learned a lot from their wisdom and life experience-but she does worry that some of the older students who will not have grown up with computers may have found the move to online classes difficult. She has particularly benefitted from how confidently AR and the peer trainers always speak about the problems they have had-not just for the learning and inspiration of listening to them, but because their openness makes her feel that she herself can say anything without being judged.
This feeling of acceptance is important to her, because she worries that her head injury might have affected her ability to remember what she has said already.
Heidi has found lockdown extremely difficult, especially because it coincided with her accident. She has always been a very sociable person, and her injury had left her unable to drive to meet people outdoors. Before lockdown, she had relied upon the sunshine abroad to help her mental state through each winter, and she struggled with not being able to travel last winter. However, RiM has really helped her to cope, because the morning sessions have helped her to get going for the day. She is writing in her journal a lot at the moment, which is something else she learned from the classes.
It was Heidi's birthday recently, and she felt hugely touched that the class remembered and wanted to celebrate it with her during the session. Now that the lockdown is easing, she is beginning to meet other students for walks and coffees, which is something she greatly enjoys. Heidi suggests that, as RiM returns to face-to-face classes, the group numbers are kept small initially, so that students do not find the transition too overwhelming.
interpersonal relationships, and in resolving mental health difficulties which had been preventing them from living their lives to the full.
Considering the comparatively poor mental wellbeing and probably high needs these students had at baseline, their own subjective perception of achieving their personal goal over the RiM course is particularly positive. Additionally, many students felt sufficiently confident to set goals that they could not necessarily easily attain.
Particularly for those older students who had been treated for many years in traditional mental health settings which focused on 'maintenance' rather than 'recovery', this level of ambition reflects a paradigm shift and an embodiment of RiM's recovery ethos.

|
The originality and value of study findings

| What this research adds
This research found RiM to benefit the mental health of its students in two ways: firstly, through its curriculum of toolkit skills; and secondly, through the sense of belonging that its 'hidden curriculum' of social learning and empowerment promotes. Whereas the educational concept of the hidden curriculum has already been identified as of value to the RC movement, 12 this study is the first to explore specifically what an RC hidden curriculum consists of and teaches, and the first to interrogate in depth how and why an RC hidden curriculum benefits and improves student mental health.
Within the RC model, peer trainers have long been regarded as key facilitators of co-production. 11,13 Whereas this study very much endorsed the co-productive value of RiM's peer trainers, this research found that student co-production is similarly important, both in improving individual mental health outcomes and in maximizing the effectiveness and accessibility of RiM itself, both before and during the Covid-19 pandemic.

| The hidden curriculum
As educational establishments rather than mental health providers, RCs aim to teach and aim to empower students through what they teach. 11 This research found that students benefitted from RiM both as a result of the 'toolkit' skills they learned from courses, and because of their less tangible experience of simply being part of RiM.
Most previous studies of RC outcomes have focused on these more measurable and more readily articulated skills. 18,20,22,38 King et al. 39 found that students often find skills-based content more accessible when delivered in an RC than within a more clinical or hierarchical mental health setting. During the Covid-19 pandemic, Rapisarda et al. 29 found that students experienced skills-based courses as more engaging than other online courses, with lower levels of student attrition also facilitating demonstrable quantitative validity.

| Co-production
Central to the RC model is the role of the peer trainers, former students who make use of their lived experience and recovery to codeliver courses and to provide support to students. 11 In this research, RiM students spoke highly of how AR and the peer trainers drew upon their knowledge and experience. Students spoke particularly highly of AR's delegation and moving forward, which meant that the peer trainers and staff encouraged an atmosphere in which students could learn from, support and empower one another.
Notwithstanding the RC model's ethos of co-production and of valuing lived experience as equal to professional experience, most RCs are led and managed by mental health professionals, and many RCs often employ their peer trainers on lower pay grades and less secure contracts than the professional colleagues with whom they share equal responsibility. 14 RiM has succeeded to some extent in addressing the power differentials between professionals and peer trainers by continuing to be led by AR identifying as a person with lived experience, and by remunerating non-NHS OTs and peer trainers in the same way as one another. However, the team remains mindful that, whereas zero-hour contracts may offer peer trainers the flexibility they need to remain employed, zero-hour contracts may also reinforce the economic marginalisation of people with mental illhealth. 44 RiM is aware progress is still needed.
This study identified co-production occurring amongst students and peer trainers whereas other RC evaluations have identified coproduction as the preserve of peer trainers, 15,45 particularly during periods of service development and change. Bester et al. 24 and Ali et al. 46 identify this as primarily due to the challenges of embedding RC activity within a host organization more accustomed to engaging co-productively with specific 'experts by experience' rather than with a cohort of service users. As an independent RC, RiM has never faced this challenge. This research found, perhaps as a consequence of this independence, that RiM students as well as peer trainers are engaged in meaningful co-production. By promoting co-production amongst students, RiM, therefore, demonstrates the benefits of close fidelity to the RC ethos of empowerment not only for the sake of involvement but as a vehicle to leadership and change. 11,13 Crowther et al. 17 and McGregor et al. 12 both assert that student co-production can be most meaningfully sustained in independent RCs.

| Strengths and limitations
Even before Covid-19, research into RCs has faced a number of unique challenges in demonstrating their validity and reliability. 18,26 Because RCs, unlike most mental health services, operate on a self-referral and opt-in basis, 11 people who participate in research into RCs tend to be those who have benefitted from them, meaning that research into RCs tends to generate positive findings. This study sought to mitigate against this bias by purposively seeking to recruit students who RiM were aware were not finding their participation unproblematic. Nevertheless, because RCs have a strong ethos of encouraging students and peer trainers to move forward with their lives, 11 they do not necessarily retain contact details for successful and satisfied former students who might generate the most positive findings.
Given the relative homogeneity of students in terms of gender, age range and ethnicitymostly White women in their 40s and 50s Researchers believe some non-attendance or attrition was partly due to interruptions in students' engagement with RiM as much as other aspects of day-to-day life throughout the pandemic. Some may be due to data collection via post or email (with phone support) instead of in-person administration.
In view of the limited unmatched sample data across time points, the planned statistical analyses became inappropriate. Instead, a descriptive approach was used to chart the changes in students' wellbeing outcomes (using WEMWBS) over time, and results should be interpreted with caution. Pandemic restrictions meant that interviews had to be rescheduled and conducted online, potentially jeopardizing the alliance between interviewers and interviewees. Given the pandemic is known to have been particularly detrimental to individuals with pre-existing mental ill-health concerns, 5,6 findings therefore should not be taken as overall RiM outcomes.

| CONCLUSION
Despite the challenges posed by Covid-19, RiM continued throughout the pandemic to uphold its 'hidden curriculum' of belongingness and co-production, and its emphasis on co-learning and co-production. Future research might consider what other RCs have learned from the Covid-19 pandemic, and how this might inform co-production to improve recovery outcomes. This research emphasises co-production as not only a tool for collective empowerment or service improvement but as a valuable skill for personal mental health recovery. In so doing, this research demonstrates RiM continued fidelity to the RC model of coproduction as crucial and integral to its success. Even when operating until the most unforeseen or challenging of conditions, 8 RCs should always endeavour to prioritize and maintain coproduction.