Understanding peer mentorship in supporting self‐management of hip and knee osteoarthritis: A qualitative study of mentees' perspectives

Abstract Background Hip and knee osteoarthritis (OA) are common musculoskeletal conditions. Treatment is usually conservative, making self‐management a priority. We developed and trialled an OA peer mentorship intervention to support self‐management in older people. Our objectives were to gain understanding of the perceived challenges of living with OA and explore how a peer mentorship intervention can support tackling these challenges; and to explore mentees' experiences of receiving the intervention to understand how this affected their OA self‐management. Methods Qualitative semi‐structured interviews focussing on acceptability and feasibility of being in the study were conducted with mentees. Transcribed interviews were double coded and subject to framework analysis. To address the objectives of this paper, three main themes were subject to focused analysis: mentees' experiences of OA, experience of peer mentorship support and factors influencing self‐management. Results Seventeen mentees participated in an interview following completion of the peer support intervention. Themes emerging from focused analysis were the following: tackling the challenges of living with OA pre‐ and post‐intervention; and the interplay of the peer mentorship intervention and self‐management. Key elements of the latter theme are enabling factors provided by peer mentorship, and mentees' readiness to self‐manage. Conclusion To effectively support OA self‐management, peer mentorship interventions should include core educational components and focus on strategies that enhance key enablers of self‐management. Paying attention to the mentor–mentee relationship and timing of intervention engagement can maximise opportunities for older people to adjust and transition from supported to independent self‐management.


| INTRODUCTION
Osteoarthritis (OA) is a long-term, age-related musculoskeletal condition affecting around 8.75 million adults in the United Kingdom (Swain et al., 2020;Versus Arthritis, 2019). OA most commonly occurs in the hip and knee joints causing pain, stiffness and functional impairment. These symptoms impact significantly on daily living and quality-of-life particularly in older adults (Bay et al., 2020;Salaffi et al., 2005). With ageing populations, hip and knee OA present a high global burden of disease (Cross et al., 2014;Swain et al., 2020;Turkiewicz et al., 2014). Evidence supports the value of selfmanagement for long-term conditions (Hughes et al., 2020;Silver, 2018;Turner et al., 2015) and the UK National Institute for Health and Care Excellence (NICE) guidance on the assessment and management of OA identifies self-management as a feasible treatment option (NICE, 2014(NICE, , 2015. The principle of self-management is that patients are enabled to become active partners in managing their health through appropriate provision of tools and techniques to address the day-to-day problems presented by their long-term condition (Foster & Fenlon, 2011). Key components of effective self-management programmes include problem solving, decision making and tailoring to individual needs (Lorig & Holman, 2003). A systematic review of chronic pain self-management (Devan et al., 2018) highlighted feeling empowered, being supported and incorporating self-management strategies in regular practice as being fundamental to success. Barriers included sustaining motivation to practice self-management and symptom-induced emotional stress.
A systematic review of musculoskeletal pain self-management interventions identified that programmes incorporating both psychological and exercise components were effective (Taylor et al., 2016) and that core elements of sustained self-management include accommodation of coping strategies into everyday life, self-acceptance and development of an identity other than 'person with pain'.
Supported self-management is an effective approach to coping with the challenges presented by OA. Successful self-management requires that individuals become motivated to adopt helpful behaviours, have belief that their efforts are effective, be alert to symptom changes and confident that they can sustain self-management beyond the supported term (Adams et al., 2017;Berry et al., 2021;Thorstensson et al., 2015). Inclusion of peers (people who share the same condition) enhances opportunities for individual relevance.
Peer mentorship is a type of self-management support where an individual with lived experience (mentor) provides support and advice to another individual with the same condition (mentee). This type of 'relational support' appears critical to self-management of chronic conditions and it 'fuels all other types of support' (Dwarswaard et al., 2016). Specific studies on peer mentorship interventions for chronic pain and inflammatory arthritis highlight the significance of personalised support and rewarding peer relationships to increase mentee engagement with self-management (Matthias et al., 2015;Sandhu et al., 2013). Timing of interventions and individual readiness to self-manage are important considerations for success (Blakemore et al., 2016;The Health Foundation, 2016).
We developed and trialled a peer mentorship intervention to support self-management in older people with OA to assess: the feasibility of an OA peer mentorship intervention; and acceptability of the intervention to peer mentors and mentees. The feasibility study and qualitative evaluation of peer mentors' perspectives are reported elsewhere (Anderson et al., in press; Lavender et al., in press). This paper provides the qualitative exploration from the mentees' perspectives.

| Aim
Our aim was to gain a clearer understanding of how the peer mentorship intervention can support OA self-management. Specific objectives were to: 1. Identify mentees' perceived challenges of living with OA and explore how the peer mentorship intervention supported them in tackling these challenges 2. Explore mentees' experience of receiving the peer mentorship intervention and its impact on their OA self-management 2 | METHODS This qualitative study used semi-structured interviews with mentees to explore their views on study participation and the peer mentorship intervention. Consolidated criteria for reporting qualitative research guided analysis and reporting (Tong et al., 2007).

| The peer mentorship intervention
Study participants receiving the intervention (mentees) were matched with a trained peer mentor who provided a programme of personalised structured support designed to aid OA selfmanagement. Peer mentors were older volunteers (>50 years) with a confirmed diagnosis of hip and/or knee OA. Their 2-day training included practical, theoretical, and socio-emotional topics around managing OA such as muscle strengthening, activity pacing, goal setting and connectedness; and incorporated mentoring skills (e.g., active listening and safe lone-working). The training programme is reported elsewhere (Anderson et al., in press). Mentees were offered up to eight, weekly 1-h support sessions in which they worked with their peer mentor to: learn about their condition and strategies for self-management (e.g., embedding activity, pacing); identify individual support needs; and develop self-management skills including LAVENDER ET AL.
-181 identifying and setting achievable goals. All mentees were shown exercise techniques and encouraged to practice muscle strengthening with their peer mentor in the session and alone. Mentees were given handouts to refer to in-between sessions. The intervention period ran between February and October 2019.

| Recruitment
Mentees were invited by letter to take part in a semi-structured interview following the end of their trial involvement. Twenty-four study participants received the intervention, 21 agreed to be contacted about participating in an interview. This formed our convenience sample. Participant information packs were posted to these mentees and interviews were arranged. Two mentees did not respond, one declined to participate and one withdrew from the interview due to time constraints.
To explore the experiences of mentees at varying time points following completion of the intervention, interviews took place at either 8, 16 or 20 weeks following baseline.

| Data collection
Semi-structured interviews were conducted in person by an experienced qualitative researcher (EDR), previously unknown to mentees.
Interviews took place at mentees' homes between August and November 2019 and followed a topic guide informed by previous literature and developed by the research team (Table 1). No subsequent changes were made to the topic guide although probes to encourage expansion of specific topics were introduced early on.
Interviews focused on key elements of feasibility and acceptability of the overall study including perceptions of the peer mentorship intervention, experience of self-management and recommendations for improvements.
Written consent was obtained prior to interviews. Mentees were assured of confidentiality and data anonymity. Interviews were digitally recorded with mentees' permission. Pseudo-anonymised recordings were independently transcribed. Returned transcripts were checked for accuracy by EDR and interview feedback was shared within the team. Topics arising during interviews were discussed and points for further exploration were highlighted. Data saturation was assumed when no new topics or perspectives arose in subsequent interviews.

| Data analysis
Interview transcripts were analysed using framework methods, allowing for team-led transparency and flexibility (Gale et al., 2013;Ritchie et al., 2003). Familiarisation and systematic open coding of the data was undertaken individually by two experienced qualitative researchers (EDR and EL). Codes were generated independently through a process of constant comparison (Bradley et al., 2007). Points of divergence were discussed and agreed. This inductive process enabled identification of emergent categories forming the basis of the analytical frameworks used for charting. Coded data were entered into frameworks independently, but sense checked by both researchers. Triangulating researcher-identified codes enhanced data credibility; validity was strengthened by repeated reflexive interrogation of the data by the researchers (Braun & Clarke, 2014;Korstjens & Moser, 2018;Spencer et al., 2013). Themes arising from the interview data were discussed with the wider research team and presented as preliminary findings at an early dissemination event attended by mentors and mentees. Feedback from this event prompted further interrogation of the data, refinement of themes and development of sub-themes.

| RESULTS
Seventeen mentees (11 women and 6 men) participated in semistructured interviews lasting on average 45.5 min (range 21-67 min). The mean age was 71.2 years (range 58-84). Notably, over half the sample had been living with OA for 5 or more years; most had OA in several joints; around two thirds lived alone, and the majority were retired ( Table 2).
Analysis of the complete interview data resulted in six themes relating to participation in the feasibility trial. To address the specific aims of this paper, the following three themes were subject to focused analysis: 1. Mentees' experience of living with OA 2. Experience of peer mentorship support 3. Factors influencing self-management Two important sub-themes emerged which aide understanding of peer mentorship support in self-management of OA: tackling the challenges of living with OA; and the interplay of the peer mentorship intervention and self-management. The results reported here focus on these two sub-themes. Mentees are referred to by pseudonyms to preserve anonymity. Duration since OA diagnosis has been included to add contextual meaning.

| Tackling the challenges of living with OA
Mentees highlighted physical, emotional, practical, financial and social challenges resulting from OA. They predominantly reported that OA-related pain and stiffness affected mobility. Walking was commonly problematic. Mentees reported challenges navigating steps, shopping, gardening and housework. Several had difficulties with dressing and disturbed sleep. Although challenges faced by mentees varied in severity, they were often pervasive and interconnected. One mentee was forced to leave her job due to debilitating OA symptoms, impacting on her socially and financially. Others feared financial consequences of being unable to work due to symptom deterioration. Typically, mentees were wary of aggravating 182 -LAVENDER ET AL.
OA by doing the 'wrong thing' and feared that everyday activity could have negative impacts long-term.
'And also fear around it thinking if I stand this long will it make it worse in the long run, so a lot of questions that I had that I didn't understand.' (Dawn, OA 1-2 years) Mentees highlighted the emotional impact of OA. They found persistent pain, stiffness and limited mobility distressing, 'And it made me want to cry, you think God, is this it?
Am I going to be so stiff, I can't even put my pants on, or my socks?' (Rachel, OA 1-2 years) Commonly, mentees were frustrated by a sense of restriction on their desire to remain physically and socially active. Those who were accustomed to variations in the severity of their symptoms approached activity pragmatically, for example, carrying a walking stick, and understood that better and worse times with OA followed a natural course.
'…it has been in remission…but then, it will flare up and be very, very painful and debilitating' (Lorna, OA 11-20 years) Several mentees reported having at least one additional health condition including cancer, chronic back pain or heart disease. Coping with co-morbidities may have resulted in the impact of OA being overshadowed for some, whereas for others the onset of OA was a significant event that prompted action.
'No, no, I was that upset, I just thought, I'll try anything. This mentee understood that she would likely benefit from becoming more physically active but lacked impetus to make a start.
In contrast, another mentee persisted with a challenging exercise regime prompted by fear of symptom deterioration.
'But I've got to a stage where I'm frightened to stop.' (Rachel, OA 1-2 years) Approximately two-thirds of mentees had previously accessed physiotherapy, privately or through GP referral, although many suggested that they found the prescribed exercises too strenuous and difficult to sustain.
'I'd done loads of going to physios and I think they give you exercises that are too difficult 'cause they're all about weights.' (Ellen, OA 11-20 years) Mentees understood that OA is not cured by exercise but some lacked belief that maintaining activity and practising exercises could improve their symptoms.
Around two-thirds of mentees suggested that they regularly used over-the-counter medication to manage OA related pain and enable them to maintain daily activities.
'I've learned the joys of paracetamol. It's a go-to drug for me now.' (Howard, OA 1-2 years) Several mentees believed that a joint replacement was inevi- In summary, mentees' pre-intervention self-management ranged from no activity to challenging daily activity; taking regular and selective pain relief; modifying activity and diet in the hope that this would bring about relief. Mentees varied in relation to their longterm and short-term approach to self-management and in their belief that OA symptoms could be modified by their efforts.

| Tackling challenges post intervention
Of the 17 mentees interviewed, twelve reported positive uptake of OA self-management behaviours due to the peer mentorship intervention; two reported making limited changes to their behaviour and three mentees reported making no changes to their self-management behaviour due to the intervention.
Mentees predominantly reported following recommendations to practise muscle strengthening exercises and increase their general physical activity for example, through walking, cycling or swimming.
Those who were physically active pre-intervention were encouraged to continue with their chosen activity. Mentees were helped to modify their ambitions to return to sport by trying a new activity (e.g., tai chi) or adopting a less strenuous alternative (e.g., walking football).
A significant change for many was introducing goals and working to achieve them. This helped mentees feel more in control of their OA symptoms, monitor their progress and sustain change for longer.
Mentees who were able to develop regular exercise routines were encouraged by the positive results.

| Understanding the interplay of the peer mentorship intervention and self-management
The majority of mentees reported benefitting from mentorship support, although not all were motivated to self-manage as a result of the intervention. Data on the challenges presented by OA, along with mentees' pre and post-intervention responses, revealed that some mentees more readily engaged with the intervention, and some more readily made changes to selfmanagement behaviours. These groups were not mutually exclusive. Central to the interplay of the intervention and selfmanagement are two key concepts: enablers of self-management and mentee readiness to self-manage. In this context, enablers were the features of peer mentorship that enhanced uptake of self-management behaviours, namely: improving understanding, building confidence and engendering motivation (Table 3).
Working with their peer mentor, mentees were able to understand the information provided and clarify its relevance to them.
They gained confidence to act on the information and were reassured that exercising correctly would not be harmful in the short-term and would help in the longer-term.
Through tailored support, mentees were helped to initiate selfmanagement behaviours, for example, practising musclestrengthening exercises, activity pacing, altering diet, engaging in outdoor activity. These initial behaviours were reinforced by peer mentor-led motivators such as encouragement to set and work towards attainable goals and exercise daily; and mentee-led motivators such as recognising changes to OA symptoms, feeling supported and gaining a sense of agency.
Supporting mentees to maintain self-management behaviours longer-term was fundamental to the success of the intervention. The role of the peer mentor was to enable mentees to initiate selfmanagement changes and to introduce strategies to foster mentees' motivation to sustain changes. However, the data suggests that translating initial changes into sustained change may be affected by mentee's readiness to self-manage and the perceived value of the mentor-mentee relationship.

| The peer mentor-mentee relationship
The quality of the peer mentor-mentee relationship affected mentees' desire to engage with the intervention. Mentees enjoyed their LAVENDER ET AL. Typically, mentees wanted to emulate or please their mentor.

T A B L E 3 Peer mentorship enablers of self-management behaviour
Even mentees who maintained a more formal relationship with their mentor reported a sense of accountability developed through regular one-to-one contact. At least initially, this drove them to demonstrate their efforts to self-manage.

| Readiness to self-manage
Mentees' readiness to self-manage was influenced by timing of their involvement in the study and the magnitude of their perceived OA challenge. For some, the invitation to participate was timely, coinciding with worsening OA or other health condition symptoms. These mentees were actively seeking change. Others appeared overwhelmed by their health concerns and welcomed support but were reluctant to initiate the changes recommended by their peer mentor during the intervention period. Some mentees were receptive to the intervention but were driven by curiosity rather than perceived need. These mentees tried out some self-management strategies before reverting to their usual behaviour.
Mentees who were ready to self-manage seemed to have accepted a longer-term attitude towards OA. They reported that mentorship support gave them direction and impetus to renew activity goals.
'It's made me more conscious to do the exercises, the physio exercises, and although I do walk and I am active, it's made me feel more strongly, that it is important to continue doing things like that.' (Lorna, OA 11-20 years)

T A B L E 3 (Continued)
Enabling

| DISCUSSION
This qualitative study set out to explore acceptability and feasibility of a peer mentorship intervention from the mentees' perspective.
The data provided insights into the challenges presented by OA, and mentees' self-management response pre-and postintervention. In general, mentees enjoyed participation, gained practically and personally from interacting with their peer mentor and found content and format of support sessions workable. Consistent with other studies (Adams et al., 2017;Berry et al., 2021;Wilcox et al., 2006), we identified specific challenges of living with OA to include: limited knowledge of the condition, lack of clarity around symptom management, difficulty maintaining desired lifestyle and fears for the future. Intervention efficacy appeared to be shaped by the complex interaction of key elements: dominance of OA challenges; how far the intervention addressed individual support needs; perceived value of the peer mentormentee relationship and mentee's pre-intervention approach to self-management.
To better understand the variable impact of the intervention on mentees' self-management, we explored which elements enabled an improvement in self-management behaviours. We identified three important factors: enhancing understanding, improving confidence and engendering motivation. We noted that when these enablers were provided through peer mentorship support, there were positive changes to self-management behaviour. However, this relationship is not linear and the impact of enablers on self-management appeared to be enhanced or restricted by mentees' readiness to self-manage and the mentor-mentee relationship.
When readiness to self-manage was high and a good relationship was established, peer mentorship enablers had most impact on selfmanagement. Readiness to self-manage is changeable and was greatest at initiation of the intervention, even for those with mild symptoms and existing self-management strategies. The opportunity to convert initial uptake of self-management behaviours into sustained changes is influenced by mentor-mentee relationship and evidence of symptom improvement (Simmonds et al., 2016). Our data suggest that the intervention had lower impact on mentees' selfmanagement for whom the enablers were less significant, for example, those who considered themselves to be knowledgeable about OA, who regularly engaged in physical activity and had strong existing social support. These mentees' readiness to implement new self-management strategies was low.  Nagelkerk et al., 2006). Participants in our study indicated that understanding OA and collaborating with their peer mentor better enabled them to self-manage. Acting to control their symptoms and recognising the benefits of engaging in tailored self-management improved confidence to continue with new strategies. Similarly, a qualitative study among patients with chronic pain/fatigue found the recognition of improved symptom management to be a facilitator in the continued implementation of activity pacing as a coping strategy (Antcliff et al., 2021). Those who found enjoyment in the mentormentee relationship were motivated to at least initiate selfmanagement behaviours. Motivation was further reinforced when the timing of the intervention coincided with higher readiness to self-manage.
The findings of this study suggest that the success of an OA selfmanagement intervention is reliant on overcoming factors that hinder engagement. Studies have highlighted physical (e.g., exercise), psychological (e.g., attitude to health), educational (e.g., knowledge of condition), social (e.g., sharing experiences) and system (e.g., accessibility of support) factors as reasons not to engage with selfmanagement (Chen & Wang, 2007;Shakibazadeh et al., 2011;Wilcox et al., 2006). Our peer mentorship intervention was designed to address each of these factors whilst also focussing on factors that enable engagement with the intervention (Kosteli et al., 2017).
The one-to-one nature of the intervention enabled specific tailoring of mentorship support to mentees' needs. However, study time constraints meant that this support was not always offered to mentees at a time when they were most receptive to it. A real-world setting would allow more flexibility around when the intervention is available. Using a readiness to change scale may be useful to identify mentees who are most likely to benefit from the intervention at a specific time. Conversely, using such a scale creates a risk of excluding some mentees with high support needs since it does not take into consideration the fluctuating nature of OA symptoms.
Although the mentees in the study who were ready to self-manage appeared to benefit most from the intervention, it is important to encourage self-management in all individuals with OA.

| Study strengths and limitations
A limitation of this study is that our sample lacked ethnic and socioeconomic diversity. However, the 17 participants showed variation in symptom range and severity, duration living with OA and response to OA challenges. Exploring self-management changes with a larger, more diverse sample would improve generalisability.
We were unable to interview participants who dropped out of the intervention or declined an interview, so our findings are at risk of positive bias and should be considered with caution. To reduce bias, interviews were conducted by a researcher unknown to mentees and transcripts were double coded.
Participants were interviewed at a minimum of 2 weeks and maximum of 16 weeks post-intervention. This was insufficient time to establish whether self-management changes instigated during the intervention were sustained, although mentees who demonstrated intrinsic motivation were potentially more likely to maintain selfmanagement behaviour longer-term. A study incorporating longer follow-up periods would be valuable.
A definitive Randomised Controlled Trial is required to provide evidence for real-world practical and financial viability and address the limitations caused by study sample, time constraints and inclusion/exclusion criteria; and will allow us to better assess mentees' readiness to self-manage OA. Definitive testing of the peer mentorship intervention would inform recommendations for wider implementation. We envisage that this intervention may require adoption by a patient organisation.

| CONCLUSIONS
This qualitative study explored participants' attitudes towards and experiences of a peer mentorship intervention to support OA selfmanagement. The majority of mentees discovered value in the intervention and the mentor-mentee relationship, and they reported self-management benefits. However, a minority reported either gaining information from the intervention or enjoying the social contact of mentor visits. Where only one of these elements was present, uptake of self-management behaviours was less apparent.
Our findings suggest that effective peer mentorship interventions must include key components and must also focus on strategies that enhance key enablers of self-management. We have demonstrated that the relational context and timing of mentorship support affects the intervention's effectiveness. We propose that paying attention to the enablers and moderators of productive peer mentorship relationships will create the best opportunity for early practical and psychological adjustment to self-management, which in turn optimises the chances of older people transitioning from supported to independent self-management.