Peer support in mental health services: where is the research taking us, and do we want to go there?

As peer workers ride over the horizon to the rescue of economically embattled mental health services we have occasion to pause, take stock and ask ourselves if the peer support that is currently emerging in state and insurance funded mental health services is really the peer support that we want people to put their own mental health on the line to offer? When we take peer support from the real world and transplant it into the contrived and constrained world that is mental health services, are we talking about the same peer support? If not, is there something that more meaningfully reflects and values the commitment that people bring to the peer worker role that we ought to be properly acknowledging and advocating for? And as researchers we might ask ‘what is our role in constructing and reconstructing peer support in mental health services?’ Might we do better for peer support?

Peer support is, of course, what we do when we recognise our shared experiences of disadvantage and distress, make an inter-personal connection on that basis, and come together to support and learn from each other. Within mental health, there is no doubting the benefits that people experience as a result of doing peer support, this journal having published studies about peer support for people with co-occurring mental health and substance abuse disorders (O’Connell, Flanagan, Delphin-Rittmon, & Davidson, 2017) and in early intervention in psychosis services (Galloway & Pistrang, 2018), peer support and the physical healthcare needs of people with mental health diagnoses (Bocking et al., 2018), mental health peer support for students (Byrom, 2018) and military veterans (Weir, Cunningham, Abraham, & Allanson-Oddy, 2017), and community-based peer support to reduce psychiatric inpatient admissions (Lawn, Smith, & Hunter, 2008). Service commissioners and providers are already sufficiently convinced, with new peer support services and peer worker roles springing up on a daily basis, in large part encouraged by austerity-driven workforce strategies in Higher Income Countries (HICs). The UK’s Stepping Forward to 2020/21 (HEE, 2017) mental health workforce plan proposes the development of new peer worker roles among 8000 new non-traditionally qualified jobs and it has been suggested that, globally, peer support represents a largely untapped resource in Lower and Middle Income Countries (LMICs) with the potential to address the ‘treatment gap’ between size of population and size of economy (Puschner, 2018). That is not to say that naked economics are the only value that mental health services ascribe to peer support. ‘Peer support enables recovery’ we are consistently told. Well-reported experiential benefits of peer support—increased hope in the future and an improved sense of empowerment (Repper & Carter, 2011)—strongly echo often-evoked conceptual framings of individual recovery (Leamy, Bird, Le Boutillier, Williams, & Slade, 2011) that in turn underpin the idea of recovery-focused services (Slade et al., 2014).

But as health researchers we normally expect the ‘does it work’ type question to be answered through randomized controlled trials (RCTs) of effectiveness, with meta-analyses of existing RCTs suggesting that evidence for the effectiveness of peer support is equivocal at best (Pitt et al., 2013; Lloyd-Evans et al., 2014). Those reviews are also at pains to point out that trials of peer support had, hitherto, failed to adequately describe what it is that peer workers do, how that is different from what other mental health workers are doing, and the mechanisms by which peer support brings about change. More recent trials have done a little better in their reporting, and an updated systematic review (King & Bender Simmons, 2018) indicates that peer workers are more likely to be delivering education or mentoring programmes than engaging in a more mutual peer support, often delivering interventions typically offered by clinicians such as case management, cognitive-behavioral therapy, and in some cases supporting medication adherence (Druss et al., 2010; Rosenblum et al., 2014). Bellamy, Schmutte, and Davidson (2017) note that recent trials that are indicative of effectiveness are likely to be evaluating, broadly speaking, some form of ‘peer supported self-management’ (see also Johnson et al., 2018), wondering, on balance, if this is where development of new peer workers roles in mental health services should be focused. King and Bender Simmons (2018) find that preference for primary outcome in peer support trials is beginning to focus in on a group of conceptually-related constructs around individual empowerment (e.g. Mahlke et al., 2017).

So here is where we might pause and take stock. Whatever is happening with peer support in practice, as researchers we seem to be starting to suggest that good peer support in mental health services might be about illness management; that the evidence is pointing towards peer support in mental health services most effectively manifesting as peer workers operating in para-clinical roles, complementing, or enacting existing clinical functions. As the focus on empowerment-type outcomes suggests, this is peer support that aims to strengthen the individual, to enable the individual to cope better with their mental health. Important of course, but reflecting an essentially medical model of mental health; that there is something wrong with you that a peer worker can support you to fix. Hence our question at the beginning of this paper; is this the full scope of what we want peer workers to be doing, the best we can ask and expect of peer support?

There is a coherent body of writing—most of it by or at least produced in partnership with people who are actively doing peer support—that suggests otherwise, arguing that the distinctiveness of peer support is attributable to a values-base grounded in naturally-occurring, real-world interactions between people supporting each other with their emotional distress (Mead & MacNeil, 2006). O’Hagan, McKee, and Priest (2009) identify three primary values: equal power relationships, reciprocal roles of helping and learning and a ‘whole of life’ rather than illness-focused approach. Trauma-informed approaches to peer support focus on what people have experienced in their wider life, rather than what might be ‘wrong with them’ as an individual (Blanch, Filson, Penney, & Cave, 2012), and in Intentional Peer Support both parties are ‘invited to learn and grow’, rather than one helping the other, with a focus on relationship and community rather than individual change alone (Mead & Filson, 2017). Furthermore, research exploring the mechanisms of peer support (Watson, 2017) tells us that people value peer support because of the opportunity it provides for normalizing, nontreatment-based relationships (Gigudu et al., 2015) and that, through those relationships, peer support works to strengthen wider connections to community (Gillard, Gibson, Holley, & Lucock, 2015a).

Of course for many individual peer workers and peer support initiatives within mental health services, that relational values base will be an important part of the work. However, there is another, equally coherent, body of research that suggests that, where peer workers are neither enabled nor supported to bring those values to their work, then peer support is diluted or eroded (Schmidt, Gill, Pratt, & Solomon, 2008; Gillard et al., 2015b) and peer workers, for their own wellbeing, retreat behind a generic support worker function. There is a well-established organisational science literature that tells us precisely that; where new workers are not enabled to bring their unique expertise and practice to their work then role adoption fails (Dierdorff & Morgenson, 2007). So it should also come as no surprise that our meta-analyses (Lloyd-Evans et al., 2014) indicate that peer workers are no better or worse than other mental health workers doing similar work. This begs the question of whether we are evaluating what peer workers actually do? We seem instead to be evaluating peer workers as a species (of mental health worker) while failing to consider whether the distinctly reciprocal, relational focus to peer support might offer something different (and effective) to mental health services.

As researchers we might ask ourselves ‘what is our role in all of this?’ Is there something about the way we do our research that is perpetuating this reductive dynamic? The RCT, as a methodology, assumes a pathology that is internal to the individual, that there is a known disease mechanism that results in a poor health state, that a treatment intervenes in and corrects that mechanism, and health improves. When we conduct trials of peer support we borrow those assumptions, and so we witness a circular logic at play. Our mental health services are underpinned by a medical understanding of poor mental health as the product of a pathologically disordered mind. Peer workers are being asked to perform para-clinical functions in support of that model. We then evaluate peer workers with a methodology designed to detect an individual treatment effect, with our balance of evidence indicating that peer support interventions that resonate with this approach are increasingly those that demonstrate effect. So the evidence base—as it coalesces, studies are replicated and meta-analyses become more precise—begins to indicate that good peer support is a peer support that focuses on fixing the individual as part and parcel of existing clinical treatment programmes. Funders of mental health services, quite correctly, will turn to this strengthened evidence for guidance on what models of peer support they should be commissioning going forward.

A strong argument can be made, of course, that peer support might be better situated outside of the constraining culture of formal mental health services (and that as researchers we might therefore better focus on evaluating those endeavours). While the commissioning of mental health services is looking increasingly to the not-for-profit sector as provider (Gillard, 2015), it is certainly the case that peer reviewed research of this more independent peer support is relatively lacking (Pistrang, Barker, & Humphreys, 2008). But it is also the case, as noted above, that introduction of peer workers into state mental health services continues apace—that for some people, mental health services will be where they experience much of their peer support—and so the argument remains that our research ought to properly question the values and purpose, as well as effects, of peer support in the statutory context.

So are we prepared to be a bit bolder and ask peer workers to do something different, to be something different? There is a question of whether a more socially focused model of peer support fits in the very clinically determined world of mental health services (Faulkner & Bassett, 2012) and whether it is reasonable also to expect peer workers to take on responsibility as agents of cultural change within mental health services (Gillard et al., 2015b). But there are compelling reasons to try. Mental health services are typically structured as a series of steps, each step embodying notions of clinical severity and mental capacity. Every step, up or down, is potentially a source of disruption to people’s lives. Connections to community, family and friends are severed time and again, and isolation imposed on people for whom isolation may be at the root of their distress. Peer support as a distinctly social intervention, at those key moments of vulnerability in people’s lives, has the potential to enable people to maintain connections to community and re-build disrupted relationships. The idea of peer support as a time-limited intervention is anathema to peer support as it naturally occurs in the real world, but as we noted above, mental health services are not the real world. If there are potentially damaging gaps in mental health services, where a distinctly social, rather than clinical intervention might make all the difference, then peer workers are perhaps uniquely placed to make that difference, to play a liminal role between people, services and community (Simpson, Oster, & Muir-Cochrane, 2018; Watson, 2017). The often dangerous step down from inpatient to community care, or specialist to primary care; the pitfalls in transition from child and adolescent to adult mental health services; struggles with early access to support with experiences of psychosis or other complex needs; the list goes on. And we are increasingly aware that understandings of peer support can be different in different cultural contexts, with community-driven approaches to peer support having the potential to play an important role in addressing persistent inequity in access to mental health care for marginalised groups within society (Faulkner & Kalathil, 2012).

To play a meaningful role, as researchers, in reimagining peer support in mental health services we need to resist the gravitational pull of the evidence base; the replication and reification of a para-clinical model of peer support as the best peer support. We need to pay attention to the values underpinning peer support in both the development of new peer worker roles, and the design and reporting of effectiveness trials—e.g. through more appropriate selection of trial outcome (King & Bender Simmons, 2018) and a values-driven approach to measuring intervention fidelity (Gillard et al., 2017)—if we are to avoid limiting the utility of our research to policy and practice. Trials should always be accompanied by high quality process evaluations that seek to better understand how peer support brings about change, and how that is distinctive from other forms of support. We need to embrace experiential knowledge, grounded in doing peer support, to develop better understandings of what peer support can be in the context of mental health services (see for example, recent experiments in peer supported Open Dialogue; Stockmann et al., 2017). This is especially the case before we start exporting peer support to LMIC where cultural understandings of mental health and community might differ from our own (Davidson & Tse 2014; Stratford et al., 2017), with efforts being made to explore reciprocal learning across countries and cultures (Baillie et al., 2015). And researchers with lived experience of peer support should be conducting these trials, not just in research assistant roles, but as research leaders controlling the research process (Boevink, Kroon, van Vugt, Delespaul, & van Os, 2016). Yes, if we do this the evidence base will remain frustratingly messy for a few years longer. But if we do this, as researchers, we can make an important contribution to ensuring that the great opportunity offered by peer support, austerity or otherwise, is not lost because we didn’t pause, take stock and ask ourselves ‘is this where we want peer support in mental health services to be going?’