|Home | About | Journals | Submit | Contact Us | Français|
Peer support is based on the belief that people who have faced, endured, and overcome adversity can offer useful support, encouragement, hope, and perhaps mentorship to others facing similar situations. While this belief is well accepted for many conditions, such as addiction, trauma, or cancer, stigma and stereotypes about mental illness have impeded attempts on the part of people in recovery to offer such supports within the mental health system. Beginning in the early 1990s with programs that deployed people with mental illness to provide conventional services such as case management, opportunities for the provision and receipt of peer support within the mental health system have proliferated rapidly across the country as part of the emerging recovery movement. This article defines peer support as a form of mental health care and reviews data from 4 randomized controlled trials, which demonstrated few differences between the outcomes of conventional care when provided by peers versus non-peers. We then consider what, if any, unique contributions can be made by virtue of a person's history of serious mental illness and recovery and review beginning efforts to identify and evaluate these potential valued-added components of care. We conclude by suggesting that peer support is still early in its development as a form of mental health service provision and encourage further exploration and evaluation of this promising, if yet unproven, practice.
They fully understand each other's language, thoughts, feelings, sorrows, signs, grips, and passwords, therefore yield to the influence of their reformed brethren much sooner than to the theorists who speak in order that they may receive applause.
D. Banks McKenzie1,2
This comment, made by the founder of the Appleton Temporary Home in Boston in the late 19th century, referred to the role that reformed alcoholics could play in the lives of people who were still actively addicted. This basic idea is not limited to addiction, of course, as the belief that people who have faced, endured, and overcome adversity can offer useful support, encouragement, hope, and perhaps mentorship to others facing similar situations has been well accepted since antiquity. The idea that this function can be performed—both adequately and to good effect—by someone with a serious mental illness, however, has only begun to be seriously considered. This is not to ignore the fact that individuals with serious mental illnesses have provided each other support informally both inside and outside of treatment settings and, further, that this support has been viewed by researchers to be of some benefit. Examples would be positive interchanges within the context of milieu or group therapy and the provision and receipt of social support in the context of psychosocial clubhouses, residential programs, or life in the community in general. Although it shares continuity with these phenomena, “peer support,” as we will discuss in this review, is something different.
With a few isolated exceptions (eg, Sullivan's hiring of people in recovery as aides on his inpatient unit in 1920s3), the notion that people with histories of serious mental illness—just like people with histories of addiction, trauma, cancer, or other serious conditions—could offer hope, support, encouragement, and even mentoring to others in similar circumstances can be traced to the early 1990s.4,5 Although these early efforts were initiated by mental health professionals, the idea itself is to be credited to the success of the mental health consumer/survivor movement. In its contemporary manifestation, this movement began in the mid 1970s as ex-patients began to gather around the country and lobby collectively for reforms in mental health care and against the discrimination associated with mental illness that they had experienced. In addition to political advocacy, the origins of this movement were in the established tradition of self-help and mutual support, a tradition that permeates American culture and has been studied extensively over the previous 25 years.6–12 Both the political successes of this movement and the rapid proliferation of self-help and mutual support groups deriving from this movement showed that people who had been diagnosed with serious mental illnesses could still play active and influential roles in their own lives, in society at large, and, finally, in the life of their peers. It was this final lesson from the consumer/survivor movement that led to the development of that form of peer support with which we are concerned in this article.
Our interest in the evidence base for this form of peer support is timely, as the last decade has seen the rapid proliferation—both across the United States and around the world—of peer-based interventions developed for, and offered to, people with serious mental illnesses. Both the President's New Freedom Commission and the Veteran Administration's new Action Agenda call for the broader dissemination of these interventions, and several states have secured Medicaid reimbursement for services provided by peers. In fact, a recent study sponsored by the U.S. Substance Abuse and Mental Health Services Administration found that groups, programs, and organizations run by and for people with serious mental illness and their families now outnumber traditional, professionally run, mental health organizations by an almost 2 to 1 ratio.13 The main question we consider is whether, or to what degree, there is adequate empirical evidence of the effectiveness of such peer-delivered interventions to support the growth and wide-scale use they are currently enjoying.
This question—despite sounding relatively straightforward—turns out to be complicated by a number of issues both intrinsic and extrinsic to the nature of peer-based interventions, as well as by the fact that peer support among people with serious mental illness is still in its early stages of development. Not the least of these complications is the fact that defining someone as a “peer”—which in this case means identifying someone as having a personal history of serious mental illness—tells us little about the person except for this 1 facet of his or her prior experience. As the idea of peer support has spread, in fact, it has become almost as common to encounter mental health professionals disclosing their own histories of mental illness as it has been to find people with histories of mental illness becoming providers of care. In the process of examining the evidence that has been collected thus far related to the use of peer-delivered interventions, we therefore will have to address the following questions:
In an earlier review of this topic, we suggested 3 broad categories of peer-delivered interventions.14 These included mutual support, participation in consumer or peer-run programs, and the use of consumers as providers of services and supports. Insofar as this review is interested primarily in evidence for the effects of services and supports provided to a person with mental illness, we will limit our review to the third category. We therefore will not be concerned with the effects of participation in self-help/mutual support, which we defined earlier as “a process by which persons voluntarily come together to help each other address common problems or shared concerns,”14p168 nor will we explore the effects of consumer-run programs that offer activities other than or in addition to peer support, which thereby lie beyond the scope of our review. Readers interested in the evidence base for these alternative vehicles for peers to be active in each other's recovery are referred to existing reviews on these topics.14–23
What, then, do we mean by “peer support”? As we depict in Figure 1 below, we suggest that peer support, while falling along a theoretical continuum, differs both from traditional self-help/mutual support groups such as GROW, Schizophrenics Anonymous, or Recovery, Inc.,23–29 as well as from consumer-run drop-in centers or businesses.30–33 In both mutual support groups and consumer-run programs, the relationships peers have with each other are thought to be reciprocal in nature; even though some peers may be viewed as more skilled or experienced than others, all participants are expected to benefit. We conceptualize peer support, in contrast, as involving 1 or more persons who have a history of mental illness and who have experienced significant improvements in their psychiatric condition offering services and/or supports to other people with serious mental illness who are considered to be not as far along in their own recovery process.
Unlike mutual support and consumer-run programs, peer support is thus defined as involving an asymmetrical—if not one-directional—relationship, with at least 1 designated service/support provider and 1 designated service/support recipient. This shift from reciprocity to receiver of care has been the source of considerable tension and debate within the consumer community, as peer staff have been viewed with varying degrees of suspicion concerning their having potentially been co-opted by the mental health system.34 The development of peer-based interventions outside of mutual support and peer-run program settings, on the other hand, has served to bring peer support closer to the mainstream of mental health practice. The presence of a review of the empirical evidence for this practice in this journal is 1 indication of this change.
Suggesting that peer support lies closer to mainstream psychiatric practice than mutual support or peer-run programs should not be taken, however, to indicate a simple equivalence. Also as depicted in Figure 1, it is important to note the ways in which conventional clinical and rehabilitative practices in mental health may—but also may not—differ from those delivered by peers. At the far left of the continuum are not only asymmetrical relationships but also relationships that are entered into intentionally—as opposed to occurring naturally—and that evolve in clinical or other formal service settings—as opposed to taking place in the broader community. Office-based psychotherapy is situated at this end of the continuum, offering the paradigmatic case of a well-structured, formally delimited, and professionally driven asymmetrical practice. On the other end of the spectrum lies friendship, which we take to be the paradigmatic case of a naturally occurring, bi-directional and mutual relationship that evolves primarily in natural community settings. Much ground lies between psychotherapy and friendship, however. We distinguish among the various activities that lie along this continuum based on who participates in them, how the people who participate identify themselves and their roles, and the locations where they are to be found, as well as on the nature of the relationship itself.
Having defined peer support as lying somewhere in the middle of this continuum, we now turn to the question of evidence for the various activities that occur in this sizable gray area.
Within the broadly defined category of peer support (indicated by the shaded trapezoid in Figure 1), let us begin with the case in which a person with a history of serious mental illness offers conventional case management services to another person with a serious mental illness; the case labeled “A” in our figure. To the degree that the services offered by this first person (eg, assistance with entitlements or housing) are no different from those offered by people who do not have a personal history of serious mental illness, it is questionable how much these services are to be considered “peer-based”—especially given the reluctance of many providers to disclose their own history of mental illness. To the degree that the person's sensitivity to experiences of hospitalization, medication, stigma, and so forth is enhanced based on his or her own experiences of disability and recovery—regardless of whether the person may wish to disclose this history—one could argue that this form of service delivery should be situated a bit farther down the continuum from case management provided by non-peers. It should be clear, though, that such a service innovation falls well short of mutual support or consumer-run programs.
Table 1 below describes the 4 studies that have been conducted to date utilizing randomized controlled trials to compare the effectiveness of conventional services provided by peers with the same services provided by non-peers (ie, people who have not acknowledged a personal history of mental illness). While these studies did not report detailed information on the actual activities of the peer and non-peer staff, descriptive studies of case management programs using activity logs and observation have shown that both peer and non-peer staff have practice patterns reflective of normal expectations for an intensive case management model, with no significant differences found with regard to the type of activity performed, the amount of client and family participation, and the location of service delivery.35 The same overlap in activities was true in our own study of supported socialization, in which both peers and non-peers offered to join adults with serious mental illness in social and recreational activities.36 As is evident in this table, 3 of these studies found no significant differences in outcomes based on whether the services or supports were provided by peers or non-peers, while the remaining study found that participants who received services delivered by peers had increased community tenure (including fewer hospitalizations).
From the perspective of feasibility, it is reassuring to know that peers can provide some of the same services and supports as non-peers with at least the same degree of effectiveness. Given the history of stigma and discrimination associated with mental illness, in this case it was important first to fail to reject the null hypothesis. With the exception of the study by Solomon and Draine, however—to which we will return below—it is unfortunately, but also entirely, unclear in these cases what, if any, value was added to the person's functioning in his or her conventional role, or in the conventional services or supports the person provided, by virtue of his or her own personal experience of disability and recovery. As more providers disclose their own histories of mental illness, we may raise this same question concerning the conventional clinical and rehabilitative services provided by people in recovery in their roles as psychiatrists, psychologists, nurses, social workers, and rehabilitation therapists. For this reason, and until we have additional evidence to the contrary, we suggest that it is best to characterize these services or supports not as “peer support” per se but as conventional services or supports that happen to be delivered by people in recovery.37
It is important to note, however, that this finding—that peers providing conventional services for the most part appear to produce the same outcomes as non-peers providing these same services—may well be an artifact of how early we are in the process of developing peer support as a resource for people with serious mental illness. At this stage in the process, we are still just beginning to find ways to identify and assess what the person brings to his or her work that is unique and based on his or her own personal experience of disability and recovery, as well as the nature of the beneficial impacts associated with these value-added components of care. It is quite possible that these effects may not be tapped by the conventional instruments that were used in these studies. While hospitalizations, functioning, and symptoms are important barometers of improvement, for example, they do not constitute all of what is involved in recovery. Measures oriented to the recovery domains identified by people living with mental illness are now being developed and may be more sensitive to the impacts of these interventions.
What, then, do we know thus far about the provision of services or supports that are uniquely based on—and therefore require—the person's own history of disability and recovery? Returning to Figure 1, an example of such a possibility lies closer to the other end of the spectrum and is indicated by the letter “B.” This second alternative represents the case in which a person with a history of serious mental illness is employed by a mental health agency to lead peer support groups for others with serious mental illnesses. Following closely on the heels of peer-delivered case management, this has been an especially appealing service for many clinical providers to add as an adjunct to their existing services in an effort to become more recovery-oriented. As a hybrid between professional-led support groups and naturally occurring self-help, this practice shares some features of each, while lacking the reciprocity that is core to mutual support. In this case, the person may be attempting to facilitate the same kinds of activities that typically occur in mutual support groups in the community, all the while being a paid employee of the clinic. In contrast to professional leaders of support groups, this person also has his or her own experience of the condition shared by all of the other members. The fact that these groups are neither fish nor fowl can generate a significant amount of tension, confusion, and concern among both group members and clinical staff, as well as for the leader him or herself.
The issues of confidentiality and boundaries offer examples of such tensions. From the perspective of the group members, the leader is a peer who is expected to uphold the value of privacy traditional to self-help/mutual support, in which nothing discussed inside of the group is to be shared with others outside of the group. But from the perspective of the agency, the expectation is that the peer leader abides by the same standards of confidentiality held by other clinical and rehabilitative staff, with necessary provisions for breaches in cases of imminent risk to self or others. This kind of divided loyalty for the peer leader represents only the tip of the iceberg of the conceptual confusions that reside in such hybrid combinations of peer and conventionally based practices.
An even more potentially contentious issue relates to opposing views of boundaries and friendship. Peer support staff may be viewed more like friends than non-peer case managers or clinical staff for several reasons. First, by virtue of drawing on their own experiences and the lessons they have learned from their own challenges and accomplishments in living with mental illness, peer staff are not only allowed but are in fact expected to disclose personal information and to share intimate stories from their own lives. Unless they are confined to service settings, peer staff also are more likely to gravitate to non-office settings such as coffee shops, restaurants, or other public spaces for meetings with group members, spaces that traditionally have been associated more with friendship than with health care. It is finally important for clinical staff to appreciate that self-help/mutual support and consumer-run programs typically do not have prohibitions against friendship between members but rather encourage the support that occurs between people to transcend what are perceived to be artificial group or program boundaries. For peer support leaders hired by clinical agencies, this often poses an uncomfortable but crucial issue. Are they able to maintain friendships they may have had with other people in recovery prior to being hired by an agency that provides services to these same people? How can they succeed in being “friendly” toward their clients without actually becoming friends with them? Regardless of its importance to the agency, is this a distinction that even makes sense to the clients? Can peer staff accept support offered to them by the people they serve? If not, then does this not move them closer to behaving and functioning like non-peer staff? If so, then when is a chat over coffee a reimbursable service and when is it just a friendly interaction? Perhaps partly because of these very sticky and largely unresolved issues, we were able to find no empirical studies examining the effectiveness or utility of this increasingly common phenomenon.
Which leads us, finally, to peer support proper, indicated in the figure by the letter “C.” In peer support, as we are defining it, a person in recovery from a serious mental illness is employed to offer services and/or supports to others with mental illnesses. As in examples A and B, this person is a paid provider of care who enters into an intentional relationship with others with the aim of assisting and supporting them in their own recovery. Like A but unlike B, it is clear in this case that the person is not a paid friend and is only to disclose personal information when it is in the client's best interest (as opposed to for the staff member's own benefit). Like B but unlike A, however, it also is clear in this case that the services and supports that the person will provide will utilize explicitly his or her own experiences of recovery. The question begged by this definition, however, is precisely what services and/or supports the person is providing. Our challenge is to identify those specific interventions people in recovery can offer that are based at least in part on their own personal history of disability and recovery such that other people who do not share this history would be unable to provide them, or at least be at a distinct disadvantage in their efforts to do so. The question marks appearing in box “C” in the figure indicate that this is an area in which there is much still to discover about what these services might entail.
To begin, the mutual support literature suggests a list of possible functions that are based on the person's shared history, including the following: acceptance, understanding, empathy, and a sense of community, thought to lead to increases in hope, autonomy, efficacy, and assumption of personal responsibility; role modeling and the provision of concrete and practical information to promote vicarious learning, modeling, and enhancement of coping and problem-solving skills; and exposure to alternative worldviews, ideologies, and contexts, which offer cognitive and environmental antidotes to the isolation, despair, and demoralization many people experience as a result of their contact with conventional mental health services.14,38,39 As useful a starting point as these concepts provide, they represent activities and potential impacts that have yet to be operationally defined, not to mention subjected to measurement. There have been a couple of attempts to begin to do so, however, which we will consider briefly as suggesting directions for future research on this topic.
A first effort to begin to tease apart the unique, active ingredients of peer support was made by Solomon and Draine40 within the context of the controlled trial described above. They hypothesized that levels of favorable working alliance and service outcomes would be higher in the peer-provided case management condition of their study as compared with usual case management services. While across conditions, regression analyses showed that working alliance favorably predicted quality of life, satisfaction with treatment, symptoms, and positive attitudes regarding medication compliance, between-condition results proved equivocal. Given our discussion above, this should not be surprising, however, as the peers in this study occupied conventional roles and provided conventional services.
In a second effort, Rowe, Jewell, Sells, and colleagues conducted a randomized controlled trial of the effects of deployment of “peer specialists” on clinical outcomes for persons with serious mental illnesses who also had histories of disengagement from treatment and violence, or the threat of violence, toward others—a population that in 38 states (not including Connecticut, where this study was conducted) would potentially be eligible for mandated or compulsory outpatient treatment.41,42 In this study, peer staff had been hired and trained specifically to utilize their personal history of serious mental illness as a means of enhancing their credibility with this “difficult-to-engage” population and as a vehicle for increasing hope, modeling adaptive problem-solving, and demonstrating the benefits of participation in treatment and rehabilitation. In addition to attending to their clients' basic needs through an extension of case management services, they enhanced their clients' access to opportunities to participate in enjoyable social and leisure activities and to occupy valued social roles above and beyond employment. To facilitate this focus on trust-building and “remoralization,” peer staff worked with caseloads of 12 or fewer clients at any given time and offered these services and supports in community settings as an extension of existing intensive case management teams.
For the controlled trial, a total of 137 participants were recruited from 4 sites and randomized to receive intensive case management with or without the benefit of a peer specialist. Data were collected through interviews conducted at baseline, 6, and 12 months' follow-up that assessed treatment relationships, service use, and motivation; provider ratings of participants' initial engagement in treatment; and record reviews that documented attendance at appointments. The primary outcome of this study was that participants who were assigned to peer specialists and who were rated as most unengaged at baseline showed significantly increased contacts with providers during the first 6 months of the study, compared with decreasing contacts for participants in the control condition over the same period. Participants in the experimental condition also reported feeling more liked, understood, and accepted by their providers than those in the control condition at 6 months, but these effects disappeared at 12 months. Participants who reported feeling more liked and understood at 6 months also had higher levels of motivation for treatment, as well as higher utilization of Alcoholics Anonymous/Narcotics Anonymous groups, at 12 months.
In terms of possible active ingredients, these findings appear to support peer providers' abilities to forge effective and stable working alliances early in the treatment process with clients typically viewed as among the most disengaged from traditional approaches to care. Consistent with earlier suggestions of Solomon and colleagues,43 these findings also suggest that differences between relationships with peer specialists and those with regular case managers may tend to surface early in the engagement process and eventually dissolve over time, as non-peer providers “catch up” in forming stronger working alliances with their clients.
While promising, the last study described represents only an initial attempt to identify the unique contributions that a personal history of serious mental illness may make to the cultivation of trusting and useful relationships for people with serious mental illness, especially among those who are disengaged from conventional services. Despite the fact that “peer support” appears to be sweeping the country, the unique ways in which persons with histories of mental illness and recovery can be useful in facilitating the engagement and recovery of others are just beginning to be explored and developed. This should not be taken as a discouraging comment to advocates of peer support, however, as much as a cautionary note about making claims that go beyond existing data. Our hope is that the current enthusiasm for peer support—which we share—will be joined with an equal degree of commitment and resources to establishing an evidence base for what precisely is involved in the process and what outcomes can be expected from a person's receipt of such services and supports.
At this point in time, we consider the state of the field to be similar to where research on psychotherapy stood prior to the introduction of manualization and other rigorous design features (eg, fidelity scales). Just as we were not sure early on in psychotherapy research what actually differentiated a Freudian clinician from a Rogerian or Behaviorist clinician in practice, the fact that a person is in recovery from his or her own serious mental illness tells us little about how he or she functions in the role of service provider. This fact also does not tell us much about the nature of the relationship this person will develop with clients based on his or her experiences. For peer support to move from being a promising practice to one that is truly evidence-based, much work will need to be devoted to developing models, manuals, training curricula, and fidelity measures that will enable us to determine what peer providers do with their own life experiences for whom, under what circumstances, and to what effects. Until then, we may at least be assured that we have no reason to believe that adding peer staff to existing mental health agencies and programs will in any way compromise the quality or effectiveness of current services, while there may still be much to be gained from the instillation of hope, role modeling, and exposure to alternative worldviews that peer staff represent by their very presence in the mental health workforce.
The VA Health Services Research and Development Service (HSR&D) provided partial funding for this article through support of Improving Care of Veterans by Using Consumers as Mental Health Providers (#IIR 02-009). Work on this article was also supported by the Department of Veterans Affairs Desert Pacific Mental Illness Research, Education, and Clinical Center (MIRECC).