Abstract
Background:
While formalised mental health peer support is on the increase, there continues to be a lack of consensus regarding the peer supporter role.
Aim:
The aim was to develop consensus on the essential components, personal costs, personal benefits, barriers and facilitators involved in providing mental health peer support.
Methods:
The Delphi method was used. In the first round, an exhaustive list of statements pertaining to peer support was generated from literature review and consultation with experienced peer supporters. In the second round, 147 UK peer supporters rated statements online or via post, and completed a questionnaire about experiences of providing peer support. Criteria for consensus were applied. Where there was uncertainty, statements were re-rated. Descriptive statistics and group comparisons were calculated. The final statements were grouped thematically.
Results:
Consensus was reached on statements pertaining to essential components (n = 67), personal benefits (n = 21), barriers (n = 1) and facilitators (n = 35). Formal peer support involves many skills and competencies. Most participants agreed that a wide range of personal benefits come with the role. Organisations may facilitate peer support through their values, actions and oversight. Approximately half of the sample worked in public services and were more likely to have concerns regarding pay and career progression.
Conclusion and Implications for Practice:
Results define the peer supporter role and add to the knowledge base about optimal conditions for it to thrive. Recommendations are made regarding role development and career progression, and future research to better understand personal costs and benefits.
Introduction
Peer support has been defined as ‘a system of giving and receiving help founded on key principles of respect, shared responsibility and mutual agreement of what is helpful’ (Mead, 2003). The formal employment of peers (i.e. people with personal experience) as providers of support/care has occurred in various settings for many years, for example, within alcohol services (Tracy & Wallace, 2016). In general mental health services, it first emerged in the United States where the ‘peer specialist’ role is now well established (Cronise, Teixeira, Rogers, & Harrington, 2016; Davidson, Chinman, Sells, & Rowe, 2006). The legitimacy of this type of formalised peer support practice is questioned by many survivors of the psychiatric system given that employment is typically in traditional psychiatric contexts and not peer-developed (e.g. Penney, 2018). Nevertheless, it is a growing paradigm in countries, including the United Kingdom, Canada, New Zealand and the Netherlands. The expansion of peer support roles in public health services has been identified as a key challenge in the development of recovery-oriented services (Repper & Carter, 2011; Sainsbury Centre for Mental Health, 2009; Shepherd, Boardman, & Burns, 2010; Shepherd, Boardman, & Slade, 2008).
Peer support is said to contribute unique benefits over traditional care, including the use of enhanced empathy (since the peer supporter has been ‘in the same shoes’ as the service user), instilling hope and role modelling recovery (Bradstreet, 2006; Davidson, Bellamy, Guy, & Miller, 2012; Gillard, Gibson, Holley, & Lucock, 2015). Evidence reviews have found that peer supporters can have outcomes equivalent to those achieved by other mental health professionals (Chinman et al., 2014; Pitt et al., 2013). Peer support is associated with increased recovery, hope and empowerment for people with ‘serious mental illness’ (Lloyd-Evans et al., 2014). Furthermore, there is growing evidence for the cost-effectiveness of peer support (Knapp et al., 2014; Trachtenberg, Parsonage, Shepherd, & Boardman, 2013). These data are influencing clinical guidance. The UK National Institute for Health and Care Excellence (NICE; 2014) has recommended peer support for people who experience psychosis; however, it has also highlighted that work is needed to develop the peer supporter role. Studies undertaken in the United States (Hamilton, Chinman, Cohen, Oberman, & Young, 2015), Canada (Asad & Chreim, 2016; Moll, Holmes, Geronimo, & Sherman, 2009), the United Kingdom (Gillard, Edwards, Gibson, Owen, & Wright, 2013) and Australia (Kemp & Henderson, 2012) suggest that a lack of clarity regarding the peer support role causes problems for peer supporters themselves, including perceived inequality in relation to non-peer colleagues and feelings of disempowerment.
Guiding principles of peer support have been developed by various bodies. For example, the International Association of Peer Supporters (iNAPS; 2011) recommend that the work of peer supporters be hopeful, open-minded, empathic, respectful, facilitating change, honest and direct, mutual and reciprocal, involve equally shared power, strengths-focused, transparent and person-driven. The Scottish Recovery Network (SRN; 2013) has developed a values framework for peer support practice comprising hope, experience, authenticity, responsibility, mutuality and empowerment. Implementing Recovery through Organisational Change (ImROC) suggests that peer support working should be mutual, reciprocal, non-directive, recovery-focused, strengths-based, inclusive, progressive and safe (Repper et al., 2013a). While there are commonalities across sets of values/principles, there are also differences. They are intended to help peer supporters maintain integrity in the role, though the extent to which they may be operationalised is unclear. This lack of consensus may pose problems for commissioners when choosing which best to adopt when developing peer support roles. There are implications for training; indeed much variety has been identified in training programmes for US peer specialists (Kaufman, Brooks, Bellinger, Steinley-Bumgarner, & Stevens-Manser, 2014). Development of consensus on the essential components of peer support has been called for so that robust research may be conducted to test its mechanisms and effects (Burke, Pyle, Machin, Varese, & Morrison, 2018). Some efforts at reaching consensus on the role have been made. Creamer and colleagues (2012) sought to develop guidance for peer support practice for those working in ‘high risk organisations’, including mental health settings (66% of the sample), the police, the military and emergency services, while Campos and colleagues (2016) aimed to develop guidance for mental health peer support practice. Both studies employed a consensus method known as the Delphi approach; however, both studies had a minority of mental health peer supporters as expert consultants, the large majority being non-peer professionals. The views of experienced peer supporters are paramount in any efforts to reach consensus about the role. The primary aim of this study was to develop consensus on the essential components of formal mental health peer support solely from the perspective of experienced peer supporters. A further aim was to develop consensus on the personal costs and benefits, and the barriers and facilitators encountered by peer supporters.
Method
The study was approved by the ethics committee of the University of Manchester, UK (Reference 16118) and the Health Research Authority of the UK National Health Service (NHS; Project ID: 215334).
Design
The Delphi methodology was used. Delphi studies apply systematic procedures to assess the level of consensus among appropriate experts and comprise a number of iterative stages or ‘rounds’ as detailed below (Jones & Hunter, 1995). An online and postal methodology enabled a large number of geographically dispersed peer supporters to take part. It was anticipated that a large number of participants would maximise the range of opinion, thereby increasing the validity of consensus.
Sampling
Eligible participants were aged 16 years and over, with personal experience of a mental health problem, who identified themselves as having provided support to another person with a mental health problem as part of a formal role within a UK organisation in the last 5 years. Data were collected between July and December 2016.
Approaches were made to organisations that were known by the researchers to employ peer support staff in public/statutory mental health services (i.e. the NHS) and in independent/voluntary organisations. Further organisations were identified through web searches. Approaches were made via email and contacts were asked to promote the study within their organisation. The social media platform Twitter was used to promote the study. Snowballing was also used; a ‘thank you’ email was sent to those who had taken part, asking the recipient to forward study details to anyone who might be eligible to participate. Participants could opt into a prize-draw to win retail vouchers to the value of GBP£50.
Procedures and analyses
Round 1
Key documents from the peer support literature were used to generate the initial list of statements (see online supplementary materials). Four experienced peer supporters were then consulted; they were separately provided with the statements and were asked to suggest additional statements and to highlight duplications and language that might be inappropriate or unclear. Following the integration of feedback, statements were arranged into a questionnaire format on webpages and on paper.
Round 2
Participants were first presented with the ‘Experience of Providing Peer Support’ (EPPS) questionnaire developed for this study (see online supplementary materials). It comprised demographic items and 23 questions about role experiences, including training, supervision, pay and role satisfaction. Participants were then presented with the statements generated in the first round. The sequence of presentation was essential components, personal costs, personal benefits, barriers and facilitators. For statements relating to the essential components, participants were asked to rate each statement on a 5-point Likert-type scale, whereby 1 = essential, 2 = important, 3 = do not know/it depends, 4 = not important and 5 = should not be included. For the statements relating to costs, benefits, barriers and facilitators, participants were asked to rate each on a 5-point Likert-type scale, whereby 1 = strongly agree, 2 = agree, 3 = neither agree nor disagree, 4 = disagree and 5 = strongly disagree. Participants were asked for their consent to be contacted at a later date to re-rate statements where consensus was not reached.
Analysis
Consensus criteria outlined by Langlands, Jorm, Kelly, and Kitchener (2008) was applied such that
Statements rated by ⩾80% participants as essential or important to peer support, or with which ⩾80% strongly agreed or agreed were automatically included; consensus was deemed to have been reached on these statements.
Statements rated as essential or important by 70% to 79% of participants or with which 70% to 79% strongly agreed or agreed went forward for re-rating in Round 3.
Any statements not meeting the above conditions were excluded.
Analyses were conducted using SPSS software, version 23 (IBM, 2015). For percentage calculations, the dividend was the total number of participants who responded to that statement. Chi-square tests were used for between-group comparisons of categorical data. Independent samples t-tests were used for ordinal and interval data where parametric assumptions were met. Mann–Whitney U tests were used for ordinal and interval data when parametric assumptions were violated. Welch’s t-test were used when differently shaped distributions meant that Mann–Whitney tests were not appropriate. Missing values were excluded pairwise. A significance level of p ⩽ .05 was used.
Round 3
Contact was made with willing participants via email or post. They were given the results of total sample ratings from Round 2 and asked to re-rate those statements on which consensus had not been reached. The same criteria and statistical procedures were applied as per the second round. No further rounds were conducted.
Grouping of statements
To summarise patterns and aid comprehension, conceptually similar statements were grouped together into ‘themes’ by the first author. These groupings were proposed to the other authors for review and feedback. A final decision on the composition and labelling of these themes was made by team consensus.
Results
Sample characteristics
Sample size and attrition
Table 1 provides an overview of characteristics for the sample in Rounds 2 and 3, including a breakdown for group comparisons. In total, 147 peer supporters took part in Round 2; 33 dropped out after completing the EPPS and 114 continued onwards to rate statements. Differences were examined between those who dropped out and those who continued. A significantly higher proportion of those who dropped out had a formal peer support qualification, χ2(1) = 8.22, p = .004; had themselves received peer support, χ2(1) = 4.81, p = .028; were unpaid for providing peer support, χ2(1) = 4.56, p = .033; and had lower levels of satisfaction with managerial supervision, t(127) = 2.17, p = .32. There was a gradual attrition of participants across the study; 79 participants (51.7%) rated all statements. In all, 69 consented to be contacted for Round 3 and 44 (63.8%) took part.
Sample characteristics at Rounds 2 and 3.
EPPS: Experience of Providing Peer Support; NHS: National Health Service; SD: standard deviation; –: none; EU: emotionally unstable; ASC: autistic spectrum condition; PTSD: post-traumatic stress disorder.
Work setting and conditions
About one-half of the sample provided peer support in public services (i.e. the NHS; n = 76, 51.7%). In all, 63 (42.9%) worked in voluntary/charitable organisations and 8 (5.4%) worked in other organisations, including the police (n = 5) and education (n = 1). In comparison to non-NHS peer supporters, a significantly higher proportion of NHS peer supporters were paid, χ2(1) = 35.32, p < .001; had longer training, χ2(5) = 27.29, p < .001; more frequent managerial, χ2(4) = 17.12, p = .002, and professional, χ2(4) = 11.08, p = .026, supervision; fewer peer colleagues, χ2(4) = 29.44, p < .001; worked as part of a clinical team, χ2(1) = 41.51, p < .001; worked more days per week, t(139.14) = 5.66, p < .001; were less satisfied with their pay, t(133.30) = 2.22, p = .28; and were less satisfied with their opportunities for career progression, t(138.21) = 4.07, p < .001.
Approximately one-fifth of the total sample (one-third of the NHS subsample) provided peer support full-time (i.e. 5 days per week). Almost one-third (44.7% of the NHS subsample) were the only peer supporter in their team. About half (80.3% of the NHS subsample) worked in a clinical team with other mental health professionals; the majority reported feeling accepted (78.3%) and valued (75.9%) by the team, though 38.6% reported negative experiences within the team.
Training and supervision
Peer supporters in the sample most commonly had 6 or more days of training, while 13.4% had received none. The majority (83.1%) were satisfied with training received. Almost three-quarters (72.1%) said they felt supported in their role. Participants most commonly received supervision on a monthly or less frequent basis. Supervision was more common and frequent in the NHS. About one-quarter of non-NHS peer supporters received no supervision. Satisfaction levels were somewhat higher for managerial supervision (74% satisfied) than for professional supervision (64.6% satisfied). Professional supervision was defined as ‘support with issues specific to the peer support role, for example, how to clarify one’s own boundaries and develop a personal account of recovery that feels safe to share’. It was provided by a variety of professionals, including team leaders/managers, occupational therapists, nurses, psychologists and social workers. Senior peer staff and those with specific peer leadership and coordination roles provided professional supervision for only a minority (20.3%).
Role satisfaction
Three-quarters (74.8%) of the total sample were satisfied in their role. Less than half (44.2%) were satisfied with opportunities for career progression; about one-quarter of NHS peer supporters (23.7%) were ‘very dissatisfied’ in this area.
Delphi rounds and consensus achieved
Figure 1 illustrates the number of statements included, excluded and re-rated at each round. A total of 225 statements were generated and reviewed by four peer consultants; their suggestions resulted in many statements being re-worded, added and excluded. In all, 211 statements were rated in Round 2. Following analysis, 104 met consensus criteria and were included; 86 were excluded and 21 went to Round 3 for re-rating. One statement was excluded in Round 3.
Statements included, excluded and re-rated at different rounds.
Thus, overall, consensus was reached on 124 statements, representing essential components (n = 67), personal benefits (n = 21), barriers (n = 1) and facilitators (n = 35) of providing formal mental health peer support in the United Kingdom. These are presented in Tables 2 to 5. To aid comprehension and discussion, tables are subdivided into ‘themes’ (the groupings of conceptually similar statements), which are presented alphabetically with constituent statements listed underneath. Consensus was not reached on any personal cost statements (see online supplementary materials for a table with all excluded statements). Further discussion of the statements and ‘themes’ is found below.
Essential components.
NHS: National Health Service.
The number of participants rating each statement in Round 2 ranged from n = 111 to 114 in the total sample and n = 62 to 63 in the NHS subsample. The total sample in Round 3 was n = 44 and the NHS subsample was n = 28.
Personal benefits.
NHS: National Health Service.
The number of participants rating each statement in Round 2 ranged from n = 99 to 102 in the total sample and n = 54 to 56 in the NHS subsample. The total sample in Round 3 was n = 44 and the NHS subsample was n = 28.
Barriers.
NHS: National Health Service.
The number of participants rating each statement in Round 2 ranged from n = 84 to 89 in the total sample and n = 47 to 50 in the NHS subsample. The total sample in Round 3 was n = 44 and the NHS subsample was n = 28.
Facilitators.
NHS: National Health Service.
The number of participants rating each statement in Round 2 ranged from n = 77 to 78 for the total sample and n = 45 to 46 for the NHS subsample. The total sample in Round 3 was n = 44 and the NHS subsample was n = 28.
NHS subsample analysis
Because differences were identified in the role characteristics of NHS peer supporters, consensus was examined separately for this subsample; data are presented in parentheses in tables. Visual inspection indicates similar levels of consensus on most statements; however, the NHS subsample reached consensus on two additional statements: ‘Lack of opportunities for career progression within the peer supporter role’ (a personal cost statement; 80% consensus) and ‘Low levels of pay for the role’ (a barrier statement; 80% consensus).
Discussion
In this study, a large and diverse sample of peer supporters were consulted, and consensus was reached on 124 statements regarding the essential components, personal benefits, barriers and facilitators involved in providing formal mental health peer support. Consensus was not reached about the personal costs involved. To the best knowledge of the authors, this is the largest survey of peer supporters in the United Kingdom published to date. In addition to consensus development, results provide a useful ‘snapshot’ of current peer support practice in the United Kingdom.
The 67 essential components elaborated in this study (Table 2) included most of the principles previously put forward (e.g. iNAPS, 2011; Repper et al., 2013a; SRN, 2013) and may be seen as confirmation of these principles in practice. Consensus on statements related to ‘empowering peers’ may be seen as consistent with evidence for improved empowerment as a result of peer support interventions (Lloyd-Evans et al., 2014). There was consensus that positive role modelling is essential. The latter has been theorised as an important change process underlying peer support (Gillard, Gibson, et al., 2015). Consensus was reached on many ‘unique aspects of being a peer’, in particular, the mutual/reciprocal nature of the peer relationship. Whether peer roles should be distinct from non-peer roles is a topic of ongoing debate (Gillard et al., 2013); these findings suggest peer supporters are keen to maintain a distinctive practice. Consensus was reached on statements describing ‘encouragement and active support’ of peers, as well as a wide range of ‘non-specific therapeutic competencies’. Proficiency in this wide range of competencies requires training, experience and guidance through supervision. Given the breadth of competencies deemed essential, it is unsurprising many would wish to stay and progress in the peer support role to develop and consolidate such skills. Equally, it is unsurprising they would wish to be appropriately paid for a role of this complexity.
A variety of essential component statements, grouped into the themes ‘acquiring and sharing knowledge’, ‘developing the practice of peer support’ and ‘promoting the values underpinning peer work’, indicate that participants see active involvement in the promotion and development of the profession as essential. These findings suggest a motivated workforce. At the same time, the ability of peer supporters to affect change within organisations is influenced by their limited power within organisations with hierarchical structures and cultures (Gillard, Holley, et al., 2015). As has been highlighted in other research, organisations play a crucial role in facilitating the success and sustainability of peer support (Byrne, Roennfeldt, O’Shea, & Macdonald, 2018). In our study, consensus was reached on 35 facilitator statements, which identify important organisational, management and supervisory supports. The list of identified facilitators presented in Table 5 may thus be useful to organisations considering the implementation of peer support, or who wish to audit their current peer support practice, in conjunction with other resources published in the field (e.g. Basset, Ryan, Repper, & Watson, 2012; Peer Worker Research Team, 2015; Repper et al., 2013b; Watson, Lambert, & Machin, 2016).
The sample reached consensus on all of the personal benefit statements proposed to them (Table 3). They agreed that providing peer support had benefits which have been grouped under themes of ‘improved well-being’, ‘making a valued contribution’, ‘material benefits’, ‘new learning’ and ‘social connectedness’. Findings are thus supportive of previous qualitative research on the positive impact of providing peer support (Bailie & Tickle, 2015; Faulkner & Basset, 2012; Moran, Russinova, Gidugu, Yim, & Sprague, 2012; Salzer & Shear, 2002; Yuen & Fossey, 2003).
It is notable that consensus was not reached for any personal cost statements and was reached in the case of only one barrier statement. Much of the qualitative literature to date has been concerned with potential personal costs to peer supporters and barriers to implementation, with views solicited from non-peers as much as peers (Doherty, Craig, Attafua, Boocock, & Jamieson-Craig, 2004; Gillard, Edwards, Gibson, Holley, & Owen, 2014; Hamilton et al., 2015). It is possible that negative perspectives have been previously overstated and this study presents a more balanced picture. Equally, given participants in this study were self-selected and given attrition during the survey, it is possible that results are biased by participants motivated to give a positive presentation of the role.
It is important to highlight that consensus cut-off criteria were high; a large proportion of peer supporters endorsed costs and barrier statements that were not included. For example, 61% agreed they ‘became “burnt out” by prioritising the desire to help peers over own health needs’. Low-to-moderate levels of endorsement of personal costs and barriers can be nevertheless considered important. These findings should still be considered by organisations to optimise implementation of peer support programmes and pre-plan ways to support peers who experience adverse effects.
There were differences in role characteristics and in the experiences of peer supporters who worked in public and non-public services. NHS peer supporters may be described as more ‘professionalised’; they were more likely to be paid, worked more hours, had more training and received more frequent supervision. Notably, they expressed dissatisfaction with pay and opportunities for career progression. Although the majority expressed satisfaction with their current role, there was also a desire to progress and a dissatisfaction with opportunities to do so. It was common in the NHS subsample not to have any peer colleagues and to receive role-specific professional supervision from non-peer professionals. These circumstances likely reflect the novelty of the role within public services (e.g. a scarcity of experienced/senior peer staff); given moderate levels of dissatisfaction with professional supervision, results suggest there is room for improvement. Supervision by non-peer staff has been described as part of a process of ‘co-optation’ into the dominant culture so that what is unique and different about peer support becomes lost (Alberta & Ploski, 2014).
Limitations
This study had some limitations. A number of participants queried whether the scope of the literature used to generate statements in Round 1 was broad enough. It was suggested that the literature may have been biased towards uncritical acceptance of the concept of mental health recovery, which is rejected by some (e.g. https://recoveryinthebin.org/). Although statements were checked by peer consultants, the authors acknowledge they could have been more reflexive about the choice of literature and professional perspectives brought to the research. This also illustrates the limitations of initial statement-generation through literature review, rather than through discussion with a diverse panel of experts, which in this instance was not feasible due to limited resources. A small number of non-NHS peer supporters reported that they found some statements difficult to answer because they pertained to more ‘clinical’ contexts; however, similarities in consensus levels for the total sample and NHS subsample suggest that inclusion of peer supporters from a variety of contexts did not bias results. Because this was a consensus-development study, there was a large number of statements to rate which was burdensome for participants. Attrition was high across the study with less data collected for the statements presented later. The sample was self-selected so it is possible they presented a positively skewed view of peer support; future studies that employ universal/random sampling techniques would mitigate against this risk. A large number of group comparisons were conducted, and therefore there is an increased risk of false-positive results. Finally, the research took place in the United Kingdom and, therefore, may not generalise to countries with different health care systems or cultures regarding mental health and peer support.
Conclusion and recommendations
This study draws upon key evidence in the field and responds to calls for the peer supporter role to be further defined. Results add to the knowledge base about how peer support can be best delivered. Experienced peer supporters identified a wide range of essential components of peer support which were grouped into themes as follows: acquiring and sharing knowledge, developing peer practice, empowering peers, encouragement and active support, meeting organisational requirements, modelling self-care and recovery, non-specific therapeutic competencies, prioritising unique aspects of being a peer, promoting peer work values, and having role clarity and supervision. Results therefore suggest that peer support demands many skills and competencies. This raises questions about how skills are developed and whether they are recognised. This may be of particular concern for peer supporters who are employed in paid roles or aspire to be. Dissatisfaction with payment and career progression in public services may reflect a mismatch between skills used and what employers currently regard as the scope of the role and as such are willing to remunerate. It may reflect the lack of organisational experience in employing peers. The peer supporters in this sample were enthusiastic about developing their practice. From a service perspective, it would appear sensible to retain staff wishing to make positive contributions. Payment that is commensurate with skills used and the provision of career development opportunities, perhaps including responsibilities for providing professional supervision to less-experienced peer supporters, would appear to be a pragmatic way forward.
Findings of this study may be used to develop fidelity checklists for peer support roles. They may be used by services planning to introduce peer support to assess whether sufficient preparation has been undertaken (e.g. education of non-peer staff about the role, clear job descriptions) and support structures are in place (e.g. employment of more than one peer supporter and appropriate ongoing supervision arrangements delivered by senior peer supporters) to create the best conditions for successful implementation.
This study provides some of the first quantitative evidence regarding the personal costs and benefits involved in providing peer support. Peer supporters reached consensus on a wide variety of benefits, including improvements in well-being, although no consensus was reached regarding personal costs. Further research in this area is needed to better understand the effects of providing peer support for peer supporters. In the linked paper in this volume, relationships are examined between experiences of providing peer support, personal costs and benefits, and psychosocial well being of the peer supporters in this sample.
