Peer Models in Mental Health for Caregivers and Families

Abstract

Peer-delivered mental health models may hold important benefits for family members, yet their prevalence, components, and outcomes are unknown. We conducted a review of peer-delivered services for families of children and adults with mental health problems. Randomized studies of interventions published between 1990-2014 were included if the intervention contained a component for family members and examined familial outcomes. Of 77 studies that were assessed for their eligibility, six met criteria. Familial components included coping and parenting skills, knowledge about mental health, and emotional support. Outcomes were uneven, although significant improvements in family functioning, knowledge about mental illness, parental concerns about their child, and parenting skills were associated with the intervention. Peer-delivered services for family members may have important benefits to family members and individuals with mental health problems; however, the research base remains thin. A research agenda to develop and examine these models is discussed.

Introduction

Peer models in the mental health field have grown over the past two decades both within the United States and internationally as a potentially low-cost, effective alternative or adjunct to services provided by professionals with mental health degrees (Davidson et al., 2012; Hoagwood et al., 2010). Peers are defined as persons with similar life experiences who also have a professional role (Hoagwood et al., 2010), and their lived experience is believed to make them best qualified to assist persons facing similar adversities (Davidson et al., 2006; Oh & Solomon, 2014). As a testament to their growth, it is estimated that approximately 10,000 adult peers are employed in the United States (Davidson et al.), and peer and consumer-delivered services currently outnumber conventional organizations employing mental health professionals at a rate of two to one (Repper & Carter, 2011).

There are key qualitative differences between peer models in the adult versus child mental health fields. For adults with mental health problems, peer programs are usually led by a peer who has also experienced a mental health problem, but who is farther along in his or her recovery. These peers typically deliver outreach and education, function as case managers, and in some cases provide therapy to adult consumers (for an example, see Eisen et al., 2012). In contrast, peers who work in the child mental health system, who are referred to as family peer advocates, family advisors or parent advisors, are caregivers of children with mental health problems who have navigated the mental health service system on behalf of their child (Acri et al., 2013a, 2013b; Hoagwood et al., 2010). They primarily function as advocates, role models, and allies to caregivers as they manage their child's mental health needs (Hoagwood et al.; Kutash et al., 2014). Regardless of whether the program is intended for an adult with mental health problems or a caregiver of a child, peer support services are unidirectional, from peer to recipient, rather than bidirectional and mutually supportive, as a way to differentiate these models from consumer-run and self-help groups (Davidson et al., 2012).

A unique, innovative peer model that is receiving growing attention embodies characteristics of both categories, in that they are intended for caregivers and relatives or children or adults with psychiatric difficulties, and may provide a range of supports depending on the needs of the family. For example, the National Alliance on Mental Illness (NAMI), which is a national organization for families of children with mental health problems (Hoagwood et al., 2010), offers peer-led programs such as Family-To-Family and the Family Support Group: These programs are delivered in a group format by one or more trained family members who provide education about mental illness, self-care strategies, information about community resources, and problem solving and communication skills (Brister et al., 2013).

There is a clear rationale for addressing the needs of family members, who in most cases provide the lion's share of support to children and adults with mental health problems. First, caregivers evidence high strain, burden and emotional distress (Addington, McCleery, & Addington, 2005; Angold et al., 1998; Bademli & Duman, 2014; Chien, 2008; Wu et al., 1999), which is attributable in large part to their caregiving role (Dixon et al., 2001). Second, caregivers are oftentimes called upon to provide a range of supports, including case management and advocacy services, financial aid and housing, yet often lack the information or skills needed to assist their family member (Dixon et al., 2001). And third, evidence suggests that when caregivers' needs are attended to, patient outcomes improve (Dixon et al., 2001).

Despite the emergence and potential value of peer models for family members, our understanding about the variations in these models, their components, and their effectiveness is relatively thin. A recent review that examined the provision of family support services provided to caregivers of children with mental health problems found in contrast to models delivered solely by mental health professionals, which were heavily represented (n=33, 66%), peer models (n=11, 22%) were unique in that they emphasized self-care, advocacy, and empowerment, yet most relied on weak research designs, such as pre/post evaluations with no control groups, which limited understanding of the impact of these models upon family-level outcomes (Hoagwood et al., 2010).

The purpose of this study is to expand upon this prior review by examining adult and child mental health programs delivered by peers that have been studied using rigorous research designs. The rationale for this endeavor is to provide a comprehensive lens about the state of the evidence regarding peer-led models in the mental health field and their potential benefits to family members. This exploration is essential given the practical needs of caregivers and family members and the potential risks to their emotional health due to caretaking demands.

Methods

A comprehensive review of peer-delivered mental health interventions was conducted. The methodological approach that we undertook was identical to those in our previous reviews (see Acri & Hoagwood; 2015; Hoagwood et al., 2012). To locate studies, the first author (M.A.) developed a list of search keywords and terms and organized them in three categories: 1) peers (e.g. peer support specialist, peer family advocate, peer advisor, peer support, community health worker), 2) mental health terms (e.g., mental, health, behavior, emotion), and, 3) research design (random or RCT). Next, search terms within each of the categories were linked with “or,” and larger categories were linked with “and” to capture studies that included at least one search term from all of the categories. We included both health and mental health terms in the keyword search as we wanted to capture the entire universe of interventions; a similar review of solely peer interventions for health is currently underway (Acri et al., in preparation).

An online search using both the PSYCinfo and Medline search systems was undertaken simultaneously through a university library system; duplicate articles were removed. Because search engines vary with respect to their search limits (e.g., different classification groupings for children and adolescents), database search limits were restricted to those found across both search engines. Hence, database search limits included year (1990–2014), and language (English).

Inclusion and Exclusion Criteria

Studies were included if they:

1. Randomly assigned participants,

2. Had been published (as journal articles, chapters, or dissertations) between 1990 and 2014,

3. Were interventions that were either solely or co-delivered by a peer,

4. Interventions were intended for caregivers and/or family members of adults or children with mental health problems, and,

5. Had published data.

Programs are excluded if they:

1. Targeted caregivers or parents of children with developmental conditions orsubstance use (e.g., intellectual disabilities, substance abuse).

2. Were peer support groups or consumer-run programs. We adopted Davidson etal.'s (2006) distinction between these programs and interventions that are peer-delivered, in that unlike consumer-run and peer support programs, peer-deliveredsupport is unidirectional (from peer to recipient) and delivered by a person who isfarther along in their or their child's management and/or improvement of theirmental health than the recipient, who is newer to the process.

3. Were primary prevention programs that did not directly address mental healthsymptoms.

In order to be comprehensive, we also reviewed the bibliographies of systematic reviews that were identified through the computer search to locate studies that may have not been captured initially. Figure 1 presents the search history and results.

Fig. 1. XXX.

Analysis

The lead author (M.A.) initially reviewed the results from the electronic search to determine their eligibility based upon the inclusion and exclusion criteria listed above. Full text articles of potential interventions were retrieved and their inclusion status was finalized. Reviews captured in the electronic search were also reviewed for additional interventions not captured previously. Next, the lead author developed a coding scheme (e.g., sample characteristics, program description and control or comparison conditions, description of the peer, and outcomes) to assess program elements across studies. Two of the co-authors (NR, CH) reviewed and coded each of the published studies that met initial eligibility. Monthly meetings were held to address questions regarding program elements. The first author (M.A.) also reviewed 40% of those studies to ensure inter-rater reliability. Any discrepancies were discussed among all of the authors until consensus was reached. The coding system was refined until no additional codes were needed.

Results

From the database search, 1,116 records were identified; after duplicates were excluded, 1,055 records were screened for eligibility. All but 59 articles were excluded from full-text review because they did not meet the inclusion criteria. Of the 59 full-text articles assessed for eligibility, 55 were excluded because of study design or because they did not focus on caregiver or family outcomes. Four studies were included from this search.

Three reviews were identified; examination of their bibliographies yielded 18 articles that were obtained; of them, 14 were excluded because they did not meet the inclusion criteria, and two articles were unable to be located. Two studies were included in this review, resulting in a grand total of six studies. Table 1 presents the results of the electronic and bibliographic searches.

Table 1. XXX.

First Author/Year	Sample Characteristics Size/Target Age/Condition	Program Description Title of Program (if provided) Whom Delivered By Description	Control/Comparison Condition	Peer Description and Role	Training Provided? Y/N If yes, by whom? Description	Supervision provided? If yes, by whom? Description	Constructs Measured	Outcomes
1. Chatterjee et al. (2011, 2014)	Two hundred eighty two (n=282) adults between 16-60 years of age diagnosed with schizophrenia and their caregivers. Mean of intervention group (36.2, SD=10.2), mean of comparison group 35.6 (SD=10.4).	Community-based care Peer-led Community-based care was a year-long program consisting of three phases (engagement, stabilization, and maintenance); community health workers conducted between 6-8 home visits during the first phase, visits every 15 days in the second phase, and monthly in the third phase. Content focused on psychoeducation about mental health and treatment, facilitation of linkages to self-help groups, and networking with agencies for family-related problems. Also eligible to receive facility based care (comparison condition).	Services as usual (facility based care), provided by a mental health practitioner.	Peers called Community Health Workers were required to have 10 or more years of education and to demonstrate good interpersonal skills. Peers provided psychoeducation and linkages to other services.	Yes Not stated. Peers were trained over the course of 6 weeks regarding content delivery and provided ongoing education.	Yes Intervention coordinators (psychiatric social workers) and Psychiatrists; intervention Coordinators. Weekly group and quarterly team meetings, onsite oversight of activities.	Individual Symptoms of schizophrenia Treatment adherence Feelings of stigma Family Attitudes/knowledge about schizophrenia Other Cost effectiveness	Individual Both groups evidenced a significant reduction in overall symptoms; at 12 months, greater reductions in disability were shown in the intervention group although there were no significant differences between groups regarding the number of participants who evidenced sx reduction of greater than 20%. Greater improvements in disability amongst the intervention group compared to the comparison group (improving 20% or more). At 12 months, individuals in the Intervention group evidenced greater reductions in all symptoms and disability subdomains; significant only in the general psychopathology and self-care subdomains. Significantly greater adherence was shown in the intervention condition. No significant differences between conditions regarding reported stigma; however, intervention group participants were more likely to report experiences of stigma/discrimination from pre to follow-up. Significant difference between groups. Family No significant effect of intervention upon knowledge/ attitudes about schizophrenia. Other Costs in the intervention group were greater than the comparison group. Supervision of peers accounted for a third of the costs.
2. Chien & Chan (2004)	Ninety-six (n=96) families of adult relatives with schizophrenia.	The mutual support group intervention Co-led by a peer and a nursing faculty member. Conducted over the course of 12 weeks, for two hours each week. The group offered information and emotional support, including cultural issues around mental health, problem solving, and caregiving demands. Individuals with schizophrenia also received treatment as usual (routine psychiatric services).	A psychoeducation group about schizophrenia and treatment. Treatment as usual (outpatient psychiatric services).	Peers were nominated by the other group members. They co-delivered the group, and facilitated the group process (e.g., establishing goals, managing needs of themselves, their relative, and coping strategies).	Yes Trained by researchers Peers participated in a two-day workshop	Yes Not stated. Reviewing session audiotapes.	Individual Functioning Hospitalizations (length and duration) Family 4. Family support (needs, service use)	Individual Compared to the other two groups, Participants in the experimental condition showed statistically greater improvements in functioning (e.g. social, living skills), and significant reductions in both the number and duration of psychiatric hospitalizations. Family No change in family support.
3. Chien & Thompson (2014)	One hundred six (n=106) families of individuals with schizophrenia. Mean caregiver age=42.7 (SD=6.0).	Family-led Peer support Group Program (FPGP) Peer led (faculty member in nursing was available as needed). Consisted of 14 sessions, each session lasting 2 hours; sessions were held every two to three weeks. The intervention was delivered over nine months. Content: offered information about schizophrenia and treatment, cultural issues, coping skills, and emotional support. Adapted from Chien & Chan's (2004) study. Also received treatment as usual (outpatient psychiatric services).	A psychoeducation group. Treatment as usual (medical consultation, nursing consultation re: medications and services, counseling by social workers).	Peers were caregivers designated by the other group members as having more experience in their caretaking role. They facilitated the group in conjunction with a researcher who guided the group in the beginning stages.	Yes Trained by researchers Peers participated in a two-day workshop	Not stated N/A N/A	Individual Functioning Hospitalizations (number and duration) Family Family support (needs, service use) Family functioning	Individual Significant improvements in the intervention group regarding functioning, number and duration of hospitalizations. No significant increase in service use. Family Intervention group participants evidenced significant improvements in family functioning.
4. Day et al. (2012) *	Caregivers of 116 children between 2 and 11 years of age with behavior problems. Mean age of caregivers 36.10 (SD=7.44).	Empowering Parents, Empowering Communities Peer led An eight-week parenting group, 2 hours each week. Content focused on parenting skills, including praise and setting boundaries.	WLC	Peers were parents; pairs of peers delivered the parenting group.	Yes Not stated The 60-hour accredited training program included practice with supervision, workshops, and the completion of a portfolio.	Yes Not stated. Biweekly supervision; supervisors also available in case of emergency.	Child Behavior Parent Concerns about child Parenting skills Parent stress Other Satisfaction with intervention	Child Significant improvements in behavior (problem and intensity) in intervention group; no change in control group Parent Significant improvements in concerns about child, parenting skills. No change in control group. No significant difference between groups in parent stress. Other Satisfaction with intervention
5. Dixon et al. (2011)	Three hundred eighteen (n=318) family members of adults who have mental illnesses. Family members were between 21 and 80 years of age, mean age of the total sample 52.2, (SD=10.6).	Family To Family (FTF) Co-led by two family members A 12-week course from The National Alliance on Mental Illness (NAMI). Group meets weekly for two to three hours. Provides information about mental illness and treatment, self care, advocacy, problem solving and communication skills, and how to enhance emotional insight about how they respond to their family member's illness	WLC- Also eligible to receive supports from NAMI or other professional services.	Peers were family members; pairs of peers delivered the group.	Yes Not stated Not stated, a side from stating that family members participating in FTF are trained to teach the program; still others receive additional training totrain other FTF facilitators	Not stated N/A N/A	Family Involvement with their relative Problem-focused and emotion-focused coping skills Mental health knowledge Burden due to their relative's illness Depression Distress (anxiety, depression, somatitization) Family functioning Use of services and supports	Family Family members who participated in FTF evidence significant improvements in problem-focused coping, and mental health knowledge than the WLC group. No significant differences between groups regarding burden. Exploratory analyses yielded additional benefits to the FTF group, including improvements in emotion-focused coping and problem solving, and reduced levels of anxiety
6. Perlick et al. (2011)	One hundred fifty-eight (n=158) caregivers of adults diagnosed with schizophrenia who endorsed feeling stigma related to their relative's illness. Caregivers in the experimental condition were 56.2 years on average (SD=11.7); control caregivers were 57.1 years on average (SD=9.7).	In Our Own Voice-FamilyCompanion Peer-led A NAMI-developed single-session program consisting of a videotape (15 minutes), and one-hour group session led by two peers. Content focused on coping with their relative's mental illness and stigma, understanding schizophrenia and how to manage the illness, and advocacy efforts.	A one-session family education program delivered by two clinicians that provided information about mental illness, treatment, its impact on the family, and community resources.	Peers were family members who either delivered or attended NAMI's Family-to-Family program.	Yes Not stated Two-day training that included guidelines around self-disclosure, techniques to facilitate the group, and strategies to discuss and characterize caregivers' feelings of stigma.	Not stated N/A N/A	Caregiver Caregiver self-stigma Anxiety Social comparison	Caregiver No significant improvement in selfstigma and secrecy (except amongst individuals in the experimental group who had low to moderate anxiety at baseline).

Sample Characteristics and Mental Health Condition

Five (83%) of the six interventions were adult mental health models. Of them, four were interventions for relatives, and one program included both the client and their family members (Chatterjee et al., 2011, 2014). Four adult mental health programs targeted schizophrenia. The fifth, Family To Family (Dixon et al., 2011), was for family members of adults with a range of mental health problems. The final program, by Day, Michelson, Thompson, Penney, and Draper (2012), was an intervention for caregivers of children with behavior problems.

Peer Characteristics

In five of the six studies, peers were family members who had lived experience caring for a relative, either an adult or child, with mental health problems. In the remaining program, by Chatterjee et al. (2011, 2014), the peer was required to have a minimum of 10 years of education and interpersonal skills; no specific mention was made of their previous or current history with mental illness.

Peer Services

Four interventions were entirely peer-delivered (Chatterjee et al., 2011, 2014; Day et al., 2012; Dixon et al., 2011; Perlick et al., 2011). These programs were 1) a year-long home-based psychoeducational intervention that provided information about mental health and treatment and facilitated linkages to community resources (Chatterjee et al.), 2) an eight-week parenting group in which peers taught parents behavioral parent training strategies (Day et al., 2012), 3) a 12-week psychoeducational and skill-building group (Dixon et al. 2011), and, 4) a one-session group intervention whereupon peers taught relatives coping skills regarding their relative's illness and stigma, how to manage schizophrenia, and advocacy efforts (Perlick et al., 2011).

The remaining two programs were team-led programs that were co-delivered by a peer and a mental health professional (Chien & Chen, 2004; Chien & Thompson, 2014). The first, a 12-week group, provided relatives with information about schizophrenia and treatment, emotional support, and problem solving skills (Chien & Chen). The second, which was an adaptation of Chien and Chen's group, was a 14-week group that offered similar content in addition to coping skills (Chien & Thompson). In both programs, the peer's role was to facilitate the group process via setting goals, assist relatives in managing both their needs and that of their relatives, and teaching coping skills.

Research Design

Two studies compared the intervention to both services as usual (e.g., outpatient mental health services) and a comparison intervention, such as a psychoeducation group (Chien & Chan, 2004; Chien & Thompson, 2014), while a third program compared the intervention to services as usual only (mental health treatment; Chatterjee et al., 2011, 2014). Of the remaining three programs, two compared the intervention to a control condition only (Day et al., 2012; Dixon et al., 2011), and one study compared the intervention to a comparison treatment (one-session family education program; Perlick et al., 2011).

Outcomes

Client

Across studies, the intervention was associated with positive outcomes including significant reductions in functioning (e.g., social and living skills), and psychiatric hospitalizations (Chien & Chan, 2004; Chien & Thompson, 2014), significant improvements in behavior (Day et al., 2012), and adherence to treatment (Chatterjee et al., 2011, 2014) in comparison to a comparison or control group. Additionally, although Chatterjee et al found both groups improved regarding symptom reduction at posttest, greater reductions in symptoms and disability were found at 12 months amongst the intervention group.

Caregiver and Family Member

Caregiver and family-level outcomes were mixed; one study found significant improvements in family functioning amongst the intervention group (Chien & Thompson, 2014), while Day et al. (2012) found significant improvements in parental concerns about their child and parenting skills in comparison to the control group.

However, Day et al. (2012) did not find significant differences regarding parental stress levels, nor did Chien and Chen (2004) find any advantage of the intervention with respect to family support. Additionally, Perlick et al. (2011) did not find a significant improvement in self-stigma and secrecy amongst caregivers, except amongst individuals who had experienced anxiety at baseline, and Chatterjee et al. (2011, 2014) found no significant impact of the intervention upon knowledge or attitudes about schizophrenia.

Finally, Dixon et al. (2011) found significant improvements in knowledge about mental illness and problem-focused coping amongst family members who received Family To Family, but no difference across groups regarding perceived burden due to their family member's mental illness.

Discussion

The purpose of this review was to present an overview of peer-delivered interventions for caregivers and relatives of children and adults with mental health problems. This review, which describes the types and components of peer programs and their impact, has important implications for the field, given relatives and caregivers are at high risk for strain, burden and distress, are often called upon to provide a range of supports such as advocacy, housing, and financial aid, and their emotional health and wellbeing is directly tied to patient outcomes. Importantly, this review extends the existing literature by providing a synthesis of programs, including their variants, components, and impact upon familial outcomes.

Several findings of this review warrant mention. First, only one study of a peer model for child mental health met criteria for inclusion. This finding is significant; it suggests that little movement has been made since previous review (e.g., Hoagwood et al., 2010), even though there has been a growing interest in peer-led models, particularly in the child mental health field. Accordingly, peer programs for families of children have either not been developed to the point that evaluations have been conducted, or have been evaluated using weaker research designs. Future research undergirded with a clear conceptual framework and rigorous designs is needed in order to move the knowledge base about peer models in child mental health forward.

A second notable finding was that out of the 59 articles that were assessed for eligibility, all but six were found to be ineligible either because they did not provide support to caregivers/family members, or because they did not examine family-level outcomes. Extending our understanding of the impact of peer models beyond the client level is necessary given the effect of mental health problems upon the entire family, and the potential benefit of peer support to both caregivers/family members as well as the individual experiencing the mental health problem.

And third, outcomes associated with the peer models were uneven; some studies showed benefits in outcomes such as family functioning, parental concerns about their child, knowledge about mental illness, and parenting skills. Other constructs such as parental stress, family support, perceived burden and coping were unchanged, however. Despite this, it is noteworthy that client outcomes were largely positive in critical areas such as symptomatology, functioning, treatment adherence, and hospitalizations. It may be that providing support to the entire family mobilized and strengthened the family unit, resulting in enhanced client outcomes in these critical domains. Future exploration is needed in order to untangle the benefits of peer support for family members upon client, caregiver, and family-level outcomes, and whether there are specific types of services or supports that would improve family-level domains for the sake of their own health and well-being. This review has important implications for practice. It suggests that at a minimum, practitioners would be prudent to assess the family's health and wellbeing, and optimally provide support to or link families to supportive services to ease the weight of their caretaking roles. As mentioned previously, NAMI has several programs that provide information about resources, skill building (e.g., communication and problem solving skills), self-care strategies, and emotional support to caregivers and relatives. Attending to the needs of family members could have a significant impact upon the family's functioning, reduce caregiver stress and strain, and enhance consumer outcomes.

This review also exposes the need for additional research, including the development of peer support models for caregivers and family members, rigorous studies of peer models for children with mental health needs and their families, and careful examination of familial outcomes and family-level mediators of client outcomes (e.g., symptom and functioning improvement as mediated by enhanced family-level factors).

Peer models in adult and child mental health systems are generating increased interest as workforce shortages become more apparent and health care changes necessitate cost savings and improved access to services. The peer workforce is becoming professionalized through credentialing programs, which are making peer support a billable service, and many states and localities are investing in peer programs (Hoagwood et al., 2010). This review contributes to our knowledge about peer models and future research directions that are needed in order to advance the peer movement and enhance the health and wellbeing of children, adult consumers, and their families.