An Innovative Model of Depression Care Delivery: Peer Mentors in Collaboration with a Mental Health Professional to Relieve Depression in Older Adults

Abstract

Objectives

Traditional mental health services are not used by a majority of older adults with depression, suggesting a need for new methods of health service delivery. We conducted a pilot study using peer mentors to deliver depression care to older adults in collaboration with a mental health professional. We evaluated the acceptability of peer mentors to older adults and examined patient experiences of the intervention.

Methods

Six peer mentors met 30 patients for 1 hour weekly for 8 weeks. A mental health professional provided an initial clinical evaluation as well as supervision and guidance to peer mentors concurrent with patient meetings. We measured depressive symptoms at baseline and after study completion, and depressive symptoms and working alliance at weekly peer-patient meetings. We also interviewed participants and peer mentors to assess their experiences of the intervention.

Results

Ninety-six percent of patients attended all eight meetings with the peer mentor and PHQ-9 scores decreased for 85% of patients. Patients formed strong, trusting relationships with peer mentors. Patients emphasized the importance of trust, of developing a strong relationship, and of the credibility and communication skills of the peer mentor. Participants described benefits such as feeling hopeful, and reported changes in attitude, behavior, and insight.

Conclusions

Use of peer mentors working in collaboration with a mental health professional is promising as a model of depression care delivery for older adults. Testing of effectiveness is needed and processes of recruitment, role definition, and supervision should be further developed.

Peer mentoring involves establishment of a relationship in which two persons share some common characteristic and experience to provide needed assistance or support. Peer support services can be useful in the care of severely mentally ill persons^1,2 as well as in chronic disease management.^3,4 Evidence on the use of peers to address emotional distress in diabetes in primary care settings is emerging,⁵ but use of peers in mental health has largely focused on individuals with severe mental illness and specialty mental health settings. Furthermore, much remains to be elucidated in terms of efficacy and process, such as selection of peer mentors, training, and supervision.⁶

Our model is grounded in collaboration. The mental health professional provides integration of clinical insight, respects the use of the peer mentor’s specific knowledge in alleviating depressive symptoms, and is a link to the clinical system. The peer mentor uses an informal but clinically informed relationship to engage the older adult. Core elements of the relationship are establishment of a strong working alliance, definition of needs and goals, and work focused with the patient on change. Peers influence patients through social learning processes, by modeling behavior, offering verbal persuasion, and changing sources of self-efficacy in an informal interpersonal context.⁷

Our approach differs from other peer studies in the following ways. The purpose of the peer mentor is to establish a therapeutic relationship that focuses on behavior change to alleviate depressive symptoms using informal skills specific to peer mentors, such as sharing of personal experience and coping skills. In contrast to our approach, lay workers in mental health interventions usually provide general social support that is not focused on behavioral change with professional involvement.^8,9 We also used peers with a history of mental illness, treatment, and recovery who identify as having a chronic mental illness, to provide services to persons from non-specialty mental health settings with mild to moderate depressive symptoms. Finally, we focus on older adults and include a large group of minority older adults in our study.

The purpose of our pilot study was 1) to assess the feasibility of peer-delivered depression care with peers who have experiential knowledge of depression; 2) to determine what peer mentors could do to engage depressed older adults and alleviate depression; and 3) to obtain patient perspectives about strengths and weaknesses of peer mentors delivering depression care with mental health professional collaboration.

METHODS

Identifying Peers and Patients

On Our Own Maryland, a mental health advocacy organization that provides training of peers in principles of recovery and social support, advertised the peer positions for the study. Interested peers were interviewed for personal history of depression, length of recovery, previous mental health training, and commitment to peer mentoring. We recruited an ethnically diverse group of four women and two men who were 50 years of age or older who had a history of depression and treatment, with more than 5 years in recovery. All had previous mental health training and volunteer or work experience, finished the training, and were matched with patients.

Individuals were eligible for the study if they were 50 years of age or older, not engaged in professional mental health care, and had clinically significant depressive symptoms as assessed with a standard questionnaire (described subsequently). Patients whose depression was not fully treated with antidepressants might benefit from a psychosocial intervention and were not excluded if they had been taking the medication for more than 3 months and were not undergoing any dosage changes. Patients were clinically assessed by a psychiatrist prior to matching and provided written informed consent to participate. Fifty-six patients were screened by telephone; 30 patients met inclusion criteria and consented to the study. Three patients dropped out after the initial assessment (two patients dropped out before meeting with a peer mentor and one dropped out after one meeting). In all, 27 patients who enrolled in the study met the peer mentor for eight meetings and completed the study.

Study Procedures

Training Peer Mentors

A geriatric psychiatrist (JJ) conducted 20 hours of training to introduce community members to the role of the peer mentor. Training focused on refining the peer mentor’s use of effective communication skills, rapport-building, active listening, sharing of experiential knowledge, and expression of empathy. Issues of cultural competence, patient confidentiality, peer roles, and patient safety were addressed. Techniques such as offering different perspectives and pursuing pleasurable activities were introduced in lay terms for peers to use informally with patients.

Clinical Evaluation

A psychiatrist (JJ) conducted a standard psychiatric clinical interview of eligible patients to obtain a clinical impression used to guide peer mentors during the intervention. Expectations regarding the relation-ship with the peer mentor were discussed and patients were asked to consider goals and problems they would like to work on.

Intervention

Peer Mentor–Patient Meetings

Peer mentors and patients met for eight 1-hour weekly meetings. The goals of the meetings were to establish a strong working alliance, identify a patient-defined problem, encourage behavior change, and facilitate connections to community and formal mental health services. At the initial visit, the peer mentor asks the patient what they would like to get out of the meetings in order to establish a patient-identified goal that they can work on during the eight meetings. If the patient does not have an identified goal, the peer mentor and the psychiatrist formulate an initial goal based on the initial clinical evaluation and peer mentor meetings (e.g., decrease social isolation, improve relationship with a family member). The goal is discussed with the patient and modified as needed based on patient feedback.

Peer Mentor–Psychiatrist Supervision Meetings

The psychiatrist (JJ) met with the peer mentors weekly for an hour for supervision and collaboration. During meetings, the peer mentor reported on patient progress, shared impressions and insights, and the mental health professional provided guidance, reinforcing and/or providing corrective feedback when necessary and continued skills development for the peer mentor. The research protocol was reviewed and approved by the institutional review board of Johns Hopkins University School of Medicine.

Measurement Strategy

Depression

Depressive symptoms were assessed with the PHQ-9¹⁰ and the Beck Depression Inventory II¹¹ at all eight peer mentor–patient visits and at pre- and post-study visits. Patients scoring 5 or more on the PHQ-9 received a clinical interview and a baseline assessment to evaluate appropriateness for the study. During the clinical assessment, a diagnosis of depression was made using the Mini-International Neuropsychiatric Interview depression module.¹²

Working Alliance and Engagement

The Working Alliance Inventory Short Form (WAI-SF) was completed by peer mentors and patients at each meeting.¹³ The range of the scores for the WAI-SF is 12–84, with higher scores indicating a stronger bond and agreement on tasks and goals. At each visit peer mentors asked patients whether they had newly taken up specialty mental health care (i.e., medications or psychotherapy). We defined patient engagement as number of meetings attended with the peer mentor.

Additional Measures

Standard questions were used to obtain demographic information, medical problems, health service use, history of counseling, and current antidepressant use. Standard assessments were used at at pre- and study visits to assess self efficacy,¹⁴ social support,¹⁵ loneliness,¹⁶ self esteem,¹⁷ coping,¹⁸ and hope.¹⁹ At the conclusion of the study, semi-structured interviews lasting from 30 to 60 minutes were conducted with patients and peer mentors, digitally recorded, transcribed, and analyzed as described in the following.

Analysis Strategy

Consistent with an explanatory sequential mixed methods design for intervention development,²⁰ we carried out pre- and post-intervention comparisons of depression and working alliance, followed by analysis of semi-structured interview data. We used the Cochran-Mantel-Haenszel method (i.e., a generalization of Friedman’s test) to test for statistical significance of the correlation between peer–patient visit and repeated measures of depression, working alliance, and other outcomes, stratifying the analysis by patient.²¹ All statistical analyses were performed using SAS, version 9.4 (SAS Institute Inc., Cary, NC), with a significance level set at α = 0.05. We carried out additional exploratory analyses of quantitative and qualitative data to understand why some patients did not respond to the peer intervention. For semi-structured interviews, we used the constant comparative method, moving iteratively between codes and text to derive themes.²² After independent coding, all themes were then discussed to achieve consensus.

RESULTS

Sample Characteristics

The mean age of study participants was 68.9 years (standard deviation: 6.6 years) and most were women and African American (Table 1). Most participants had a history of a major depressive episode, half had experience with counseling, and one-third were taking an antidepressant.

TABLE 1.

Sociodemographic Characteristics of 27 Patients in the Peer Mentor Study (2014–2015)

Characteristics	N (%)
Age in years (mean ± standard deviation)	68.9 ± 6.6
Women	23 (85)
Ethnicity
Black or African American	20 (74)
White	6 (22)
Other	1 (4)
Education
Grade 1–8	2 (7)
1–4 years of high school or GED	8 (30)
1–3 years of community college or technical school	11 (41)
4 years of college or more	6 (22)
Marital status
Married	7 (26)
Divorce/separated/widowed	18 (67)
Not married but living with partner	2 (7)
Total medical problems
0	1 (4)
1–2	8 (30)
≥3	18 (67)
History of major depressive episode	22 (81)
History of counseling	14 (52)
On antidepressants currently	8 (30)

Notes: Numbers in parentheses are row percentages except for age, for which mean and standard deviation are provided.

Pre- and Post-Intervention Assessments

Depression and Working Alliance

Twenty-four patients met criteria for major depressive episode and three met criteria for minor depressive episode. Patients had significantly reduced depression over time as assessed both by the PHQ-9 (plotted in Figure 1) and the Beck Depression Inventory II (Table 2). Four participants had depression scores that remained relatively unchanged (top right of Figure 1). Initial ratings of working alliance obtained from patients were significantly higher for patients whose depression improved (bottom half of Table 3, mean WAI-SF 64.0 versus 55.1, χ² = 4.49, df = 1, p < 0.04). Working alliance as rated by both peers and patients significantly increased over time among patients whose depression improved (left-hand side of Table 3), but did not change appreciably among patients whose depression did not improve.

FIGURE 1.

Depression scores based on the Patient Health Questionnaire-9 at baseline, at meetings 1–8 (M1–M8), and at post-assessment visit (POST) for 27 participants from the Peer Mentor Study (2014–2015). Dashed lines connect through missing points.

TABLE 2.

Depression and Other Measures Assessed Before Meeting with the Peer Mentor and after 8 Weeks Of Meetings, for 27 Patients in the Peer Mentor Study, 2014–2015

Measures	Before Intervention	After Intervention	χ2	pa
PHQ-9 score	14.7 ± 4.4	7.1 ± 7.1	21.16	<0.001
Median [IQR]	14 [12, 18]	6 [2, 10]
BDI score	25.7 ± 9.5	12.9 ± 12.6	13.5	<0.001
Median [IQR]	25 [18, 33]	10 [3, 21]
General Self-efficacy score	28.4 ± 5.3	30.2 ± 5.4	1.96	0.16
Medical Outcomes Social Support score	53.1 ± 19.5	54.5 ± 21.1	2.67	0.10
Emotional/Informational Support	20.8 ± 8.3	22.6 ± 8.5	1.96	0.16
Tangible Support	12.1 ± 5.5	11.8 ± 5.8	0.2	0.65
Affectionate Support	9.2 ± 4.5	8.7 ± 4.4	0.00	1.00
Positive Social Interaction	7.8 ± 3.8	8.5 ± 3.8	1.47	0.23
UCLA Loneliness score	36.7 ± 15.4	31.4 ± 16.7	4.84	0.03
Rosenberg Self Esteem score	19.1 ± 5.1	18.7 ± 5.3	0.00	1.00
Brief COPE score	65.5 ± 13.2	69.9 ± 10.4	7.54	0.006
Adult State Hope score	27.5 ± 9.5	35.0 ± 8.8	3.24	0.07

Notes: Median and interquartile range of depression scores are provided for comparison due to the small sample size.

Values are means ± standard deviations unless noted.

BDI: Beck Depression Inventory; IQR: Interquartile range; PHQ: Patient Health Questionnaire.

^aThe statistical significance of individual variables was assessed with the Cochran-Mantel-Haenszel χ² tests with df = 1.

TABLE 3.

Initial and Final Total Working Alliance Scores And Subscales (Bond, Agreement on Task/Goal), for 23 Patients Whose Depression Scores Improved and for 4 Patients Who Did Not Appear to Improve

	PHQ-9 Scores Decreased (N = 23)				PHQ-9 Scores No Change or Increased (N = 4)
Peer Ratings	Initial	Final	χ2	pa	Initial	Final	χ2	pa
Total WAI-SF score	57.5 ± 9.2	65.3 ± 5.4	42.45	<0.001	56.8 ± 11.3	55.5 ± 10.7	9.73	0.20
Median [IQR]	58 [51, 66]	66 [62, 69]			58.5 [47.5, 66]	52 [48.5, 62.5]
Bond subscale	22.2 ± 4.3	25.5 ± 2.6	36.36	<0.001	23.3 ± 4.6	23.3 ± 3.4	4.35	0.74
Median [IQR]	23 [19, 26]	26 [24, 28]			23.5 [19.5, 27]	22.5 [21, 25.5]
Agreement on task/goal subscale	35.3 ± 5.6	39.8 ± 3.2	23.34	0.002	33.5 ± 6.8	32.3 ± 7.6	13.52	0.06
Median [IQR]	36 [31, 39]	40 [38, 42]			35 [28, 39]	30.5 [27.5, 37]
Patient ratings
Total WAI-SF score	64.0 ± 7.8	70.0 ± 6.0	22.60	0.002	55.1 ± 7.3	60.0 ± 12.8	8.74	0.27
Median [IQR]	66 [55.6, 71]	71 [67, 72]			55.5 [49, 61.2]	60.5 [49, 71]
Bond subscale	23.5 ± 3.6	26.4 ± 1.9	33.83	<0.001	20.0 ± 3.9	22.8 ± 6.2	11.75	0.11
Median [IQR]	23 [21, 28]	27 [25, 28]			20.5 [17, 23]	23.5 [17.5, 28]
Agreement on task/goal subscale	40.4 ± 5.6	43.6 ± 4.8	15.71	0.028	35.1 ± 3.9	37.3 ± 6.7	9.57	0.21
Median [IQR]	42 [37, 45]	44 [41.1, 44.6]			34.3 [32, 38.3]	37 [31.5, 43]

Notes: Median and interquartile range are provided for comparison because of the small sample size.

Values are means ± standard deviations unless noted.

PHQ: Patient Health Questionnaire; WAI-SF: Working Alliance Inventory—Short Form; IQR: interquartile range.

^aThe statistical significance of individual variables was assessed with the Cochran-Mantel-Haenszel χ² tests with df = 7.

Use of Mental Health Services

No participants reported any previous experience with peer mentors or lay counseling of any kind. Two patients were referred to specialty mental health services during meetings with the peer mentor and one patient was referred post-study, either because there was a clinical need for medications or the patient desired professional counseling. These two patients desired to continue meetings with the peer mentor simultaneously with specialty mental health care. At study completion, one African American man desired a referral to counseling and was linked to services. Twelve patients with clinically significant depressive symptoms (PHQ-9 > 5) post-study declined referral.

Patient Perspectives

Themes from interviews related to the patient’s relationship with the peer and to what the patients said they “got out” of the relationship (Table 4).

TABLE 4.

Themes and Example Quotes about the Intervention in the Peer Mentor Study, 2014–2015

The relationship with the peer
Similar life experiences: ” They’ve been there”	… a lot of people don’t understand depression … They think it’s something that you can treat and it’ll be over with. But she [peer] understood that. She understood that because she’d been through depression herself, and that made it easier for me to talk to her. (Patient 1)
The peer mentor is not a friend or a neighbor	… she accepted me how I am, and I know that if she had come off strong she couldn’t accomplish too much for me, because I’m going to shut down at some point because I feel that I have to defend myself, and I didn’t think I had to do that. (Patient 4) She kept it …. professional in that she knew when to come forward with certain things and when to kind of let me just flow. (Patient 5)
Mental health professional involvement lends credibility	… patients very much liked my having an expert in the background that I could confer with. I think it gave them another level of trust in the process. (Peer 1)
What patients said they “got out of it”
Hope and motivation	She gave me hope, she made me feel there was still hope for me, you know, not to give up, and every problem I presented to her she made me realize, ”You can overcome that, that’s no big deal,” you know … (Patient 7)
Skills Insight	Suggestions and coping skills:”I had suggestions or directions to go on”(Patient 9) Basically, as far as the relationship that I was in, she was able to help me to see some things that I had overlooked. And it’s like, sometimes you can be walking around with blinders on. But I don’t know, somebody else can see it, but you don’t see it (Patient 8) I have these tapes playing in my head…. that you always have to prove that you can do the undoable … You have to be perfect…. It was when she gave me the insight to what I was talking about that was what really helped me. (Patient 6)
Participants whose depression did not improve
Length of intervention	She [the peer] seemed to be caring but then I’m distrustful of people … there were things that I wanted to say and I wouldn’t. You’ve got to build up confidence…. And eight weeks is not long enough for me to really build confidence with someone. (Patient 13)

The Relationship with the Peer

Similar life experiences: “They’ve been there”

Patients stressed the importance of similar experiences in leading them to believe in the credibility and value of the coping skills suggested by the peer mentor. Participants valued talking to peers who “are more empathetic, because they’ve been there” and who “didn’t judge me or anything.” Peer mentors with experiential knowledge could understand and facilitate trust and disclosure.

For many participants, similarities of life experience such as being a single parent, sexual trauma, domestic violence, homelessness, or having extended families with illegitimate children were more important than the peer mentor’s experience with depression, particularly in patients who did not use the term “depressed” to describe themselves. Some participants said they enjoyed meetings with the peer because depression was not the focus.

An African American man appreciated speaking with a peer mentor who was in her 70 s because “you talk to young people, they don’t know too much about nothing no way. And they just laughing at you when you talk to them.” He also had lifelong concerns about his illegitimate birth and subsequent standing in his family. The peer mentor had similar experiences within her own large, multi-ethnic, extended family, which enabled her to understand and change his perspective about his problem to “release a lot of this anxiety” and change “how you think all your life the same thing.” An African American woman talked to the peer about sexual trauma and felt this was the main benefit she derived from meetings with the peer. The patient stated that if the peer mentor had not had personal experience of molestation himself and disclosed it to the patient, “We probably never would have got on that topic.”

The peer mentors’ ability to disclose information about their own difficult life experiences was perceived as putting the peer on the same “level” as the patient, and this facilitated disclosure and built trust. An African American woman said, “the ability to share tragedies or different situations in our life…. I think the sharing was a very, very important part. That built up a trust factor that I could talk to her.”

Matching based on ethnicity was important for some patients because sharing an ethnic background meant similarity in experiences: “By being black women we had a lot of things in common that we could talk other than about myself.” However, another felt that she could have been paired with “a person of African descent and it wouldn’t have worked.”

The peer mentor is not a friend or a neighbor

Although patients were willing to try meeting with the peer, several patients were skeptical, they “didn’t really expect too much at first,” and some were surprised they experienced any benefits. One man said that although he was skeptical in the beginning, he “found that those things [coping skills provided by the peer mentor] would work and actually take away some of the bad feelings that I had.”

Patients acknowledged “sometimes talking to some people makes you feel worse.” In contrast, the participants recognized the ways in which peer mentors showed good communication and relationship skills. Peer mentors did not impose solutions upon patients. Peer mentors were perceived to have a therapeutic attitude–focusing one’s energy on improving the patient’s well-being with self-benefit.

Mental health professional involvement lends credibility

Patients were comforted by the involvement of a professional who provided an initial clinical evaluation prior to matching with the peer and peer supervision. A peer mentor said that “patients very much liked my having an expert in the background that I could confer with. I think it gave them another level of trust in the process.”

One woman who asked for referral to a professional therapist thought the peer mentor could complement the therapist as someone who could “open the door” and help engagement with professional care. One African American woman experienced significant relief from depressive symptoms, but she did not have health insurance to obtain professional care. She commented that the peer mentor who identified “patterns” in her life was helpful and “now would be the time, then, to have more professional care and really see the areas … and how to let go of it permanently.”

What Patients Said They “Got Out of It”

Hope and motivation, coping skills, and insight

Peer mentors gave hope and motivation by sharing their stories of struggle and recovery (Table 4). Patients valued suggestions and coping skills—direction and a “practical” orientation with “offered suggestions” based on personal experience.

Talking to peers enabled the patients to “see” things and encouraged healthy behaviors. Even though the patients themselves may have realized certain things that they should think or do, the peer mentor’s encouragement enabled them to act on those ideas. Peers also helped patients gain insight leading to different behaviors, see positive things about themselves, and increased their self-esteem. One African American woman with poor self-esteem reinforced by her social environment had become discouraged. The peer mentor successfully encouraged her to take up an activity which she had given up, boosting her self-esteem and sense of empowerment.

Participants Whose Depression Did Not Improve

Our goal here was to learn what we could about four patients whose depression did not improve and who showed no significant change on working alliance scores (top right of Figure 1 and Table 3). Themes raised by the four patients who did not improve were related to difficulty with trust, perception that the 8-week length of the intervention was too short, and concerns about disclosing personal information to a non-professional. In contrast, patients whose depression scores improved “desperately wanted it,” were open to learning, and tried to apply what they learned outside of meetings.

DISCUSSION

Older adults who engaged with the peer mentor experienced decreased depressive symptoms, and high engagement in terms of attendance of all eight meetings was robust. The importance of establishing an effective working alliance and the contribution of patient factors such as ability to trust and disclose openly was clear. Patients who were engaged benefited in multiple ways, from being heard, attitude change, behavioral change, and insight. Peer mentors were able to provide an “experience” rather than “treatment,” relying less on a medical diagnosis with its associated stigma and implied judgment. The possibility of using peer mentors to direct patients to professional mental health services may be useful, but most patients in our study did not express a desire to be referred. Potential mediating factors were decreased loneliness, improved coping skills, and working alliance. Working alliance in terms of the ability to connect emotionally, define goals, and work collaboratively was stronger for patients whose PHQ-9 scores decreased during the intervention. Professional involvement did not detract from peer-delivered services; rather, involvement of a mental health professional lent credibility to the program and reminded patients that a certain level of quality and accountability could be expected.

Before discussing implications, we need to consider limitations. First, our study was not a controlled randomized trial so that resources could focus on development of the intervention. We have provided evidence, however, that patients whose depression improved also showed positive change in working alliance and described their involvement with the peer as more engaging. Second, the six peer mentors may not represent the diversity and skill of all peer mentors. Third, for this pilot study we did not strive to obtain a representative sample of patients because our goal was to develop the intervention using peers. Fourth, the sample size precluded taking into account the nesting of patients within peer mentors in the statistical analysis. In any case, the purpose of this intervention development pilot study was to assess the feasibility of peer-delivered depression care and to obtain patient perspectives in order to further develop the model.

We were encouraged to find depression improved among patients who engaged with peer mentors. Older adults do not engage strongly with current depression treatments such as psychotherapy or pharmacotherapy.^23,24 Enhanced referral to clinical services has not been shown to engage older adults significantly, suggesting that delivery of current depression treatments need to be changed for those who underutilize services.²⁵ Patients valued the informality of the peer mentor, whom they often viewed as a “friend.” This perception helped patients engage easily but, importantly, the relationship was focused on therapeutic value for the patient, and this sustained engagement beyond the first meeting.

The potential benefits of peer mentoring are significant but good skills and supervision to use those skills appropriately by the peer are needed. Peer mentors can be neutral in benefit or, at the worst, harmful.²⁶ Equally important is that patients are discriminating and may be skeptical of engaging with peer mentors. For this reason, it is important to recognize the qualities that peer mentors should have, including strong relationship-building skills, a “therapeutic attitude” in which they understand that the person receiving the help is always the focus, and an ability to give suggestions and share coping skills to encourage change. Peer-delivered mental health services in which supportive relationships are created require a complex understanding of relationship-building and behavior change, a role that a mental health professional can fill. Professionals and peers working together can be complementary and more effective than either alone.²⁷

A peer approach may be particularly appropriate for engaging hard-to-reach individuals who do not accept traditional mental health services. Socioeconomic differences and poor communication can loom large between health providers and patients, resulting in distrust of professionals.²⁸ Use of peer mentors who are from a similar socioeconomic class or ethnic community have contextually relevant knowledge and ways of approaching the minority older adult that can overcome distrust.²⁹ Our study suggests the feasibility and acceptability of a peer-based model of depression for minority older adults.

Federal insurance programs pay for services provided by certified peers and patients who meet qualifications, and some states in the United States have developed peer recovery support specialist certification processes intended to meet federal requirements for reimbursement. The focus of such programs is now largely on development of services for persons with severe mental illness and substance abuse. Nonetheless, the potential for peer mentors to provide services to depressed, minority and underserved older adults has high public health significance. Persistent disparities in mental health care for minority older adults, and the coming “tsunami” of older persons³⁰ demands that novel delivery models be developed and tested.