Table 1.

Summary of findings

Reference, country	Study design and setting	Peer intervention	Intervention and control group	Summary of findings
Disorder: serious mental illness (SMI)
Group interventions
Cook [19], USA	RCT (individual); community	Illness self-management intervention (WRAP) delivered in 8 weekly session of 2.5 h. 5–12 participants per group	Intervention: WRAP + TAU (outpatient community mental health care) (n = 276). Control: WRAP waitlist + TAU (n = 279)	Over time, greater symptom reduction (BSI: −0.05, p = 0.023), significant improvements in quality of life related to the environment (WHO-QOL: 0.39, p = 0.001) and hopefulness (HS: 0.40, p = 0.018) compared to the control group
Cook [20], USA	RCT (individual); community	Recovery education intervention (BRIDGES) delivered in 8 weekly sessions of 2.5 h. 4–13 participants per group	Intervention: BRIDGES + TAU (outpatient community mental health care) (n = 212) control: BRIDGES waitlist + TAU (n = 216)	Individuals participating in BRIDGES showed significantly greater improvement than controls in self-perceived recovery (RAS total: 1.55, p = 0.013) and in some aspects of hopefulness (SHS agency: 0.33, p = 0.006) but not in the total hope score (SHS total: 0.20, p = 0.347)
Druss [21], USA	RCT (individual); community	Health and Recovery Peer Programme (HARP): disease self-management programme. Participants met in 6 weekly sessions, 8 participants per group	Intervention: HARP only (n = 41). Control: professional health care: outpatient community mental health care provided by health professionals (n = 39) (equivalence trial)	Patients in HARP scored higher than the control group on physical (42.9 ± 14.2 vs. 40.0 ± 13.7) as well as mental HRQoL (36.8 ± 10.0 vs. 36.8 ± 11.1) at 6 months follow-up, but group*time interaction is not significant
van Gestel-Timmermans [22], Netherlands	RCT (individual); community	“Recovery is up to you” programme (structured programme consisting of psycho-education, illness management, learning from other’s experiences, social support, homework assignments) provided in a group setting over 12 weekly two hour sessions	Intervention: “Recovery is up to you” + TAU (continuation of usual treatment regimes, no further specification) (n = 168). Control: “Recovery is up to you” waitlist + TAU (n = 165)	The intervention had a significant and positive effect on empowerment (χ 2 = 10.42, p = 0.015), hope (χ 2 = 15.57, p = 0.001) and self-efficacy (χ 2 = 11.46, p = 0.009), but not on quality of life (χ 2 = 2.66, p = 0.45) and loneliness (χ 2 = 4.81, p = 0.19)
Individual interventions
Davidson [23], USA	RCT (individual); community	The Partnership Project: patients spent individual time with peer (2–4 h per week for a period of 9 months) and participated with the peer in social or recreational activities in the community promoting recovery	Intervention: (1) partnership project with a volunteer who had a personal history of psychiatric disability + TAU (outpatient care received at state-run community mental health centres) (n = 95); (2) partnership project with a volunteer partner who had no history of psychiatric disability + TAU (n = 95). Control: not matched with a volunteer partner (participated in social activities alone) + TAU (n = 70)	Differences between groups on clinical and psychosocial outcomes were only found when participant’s frequency of contact was considered: Participants in the volunteer group improved in terms of social functioning and self-esteem when meeting regularly with their partners (F(2,44) = 2.95, p = 0.06), those assigned to the peer group only improved when they did not meet regularly (F(2,214) = 3.73, p < 0.05)
Forchuk [16], Canada	RCT (cluster), community	Transitional discharge model (TDM): peers assisted individuals hospitalised with chronic mental illness in successful community living, taught community living skills, provided understanding and promoted friendship (duration of programme: 1 year)	Intervention: peer support + TDM + TAU (traditional community mental health care) (n = 201). Control: TAU (n = 189)	Quality of life of the intervention group was not significantly improved compared with the control group, F(1,22) = 0.38, p = 0.27)
Greenfield [24], USA	RCT (individual); community	Crisis residential programme (CRP) managed by peers in which consumers acted as case managers for patients (emphasising client decisions and recovery). Minimum length of stay in CRP was 8 days (max 30 days)	Intervention: peer managed CRP + TAU (treatment by a psychiatrist for medication) (N = 196). Control: TAU (treatment by a psychiatrist in locked inpatient psychiatric facility) (n = 197)	Significantly greater improvement in psychiatric symptoms in the CRP compared to control group (BPRS: −0.11, p = 0.002). Gains were also seen in both groups in GAF and QoLI, but group by time interactions were not significant (p > 0.05). Self-esteem improvements were only seen in the peer-based intervention (RSES: 0.64, p < 0.05)
Rivera [25], USA	RCT (individual); community	Peer-assisted case management over 12 months. Peers engaged clients in social activities, developed supportive social networks among clients, and contributed to treatment planning	Intervention: peer-assisted case management (n = 70). Control: TAU (strength-based intensive case management without the peer enhancement) (n = 66)	No significant differences between the peer-assisted group and the control arm in clinical or psychosocial outcome measures at 6 or 12 months (results from analyses of covariances and p value not reported)
Sells [26], USA	RCT (individual); community	Individual case management provided by peers over 12 months	Intervention: peer case management + TAU (continued other treatment received before; no further specification). Control: case management with traditional providers + TAU (n = 69)	There is no statistically difference in quality of life between the intervention and the control group at 12 months (t tests and p values not reported)
Solomon [27], USA	RCT (individual); community	Individual case management according to the assertive community treatment model provided over 12 months. Goals were determined with the client and included psychiatric treatment, social and family relations, living situation and income	Intervention: consumer case management by peers (n = 48). Control: case management by health professional) (n = 48; equivalence trial)	No significant differences between the conditions in symptom observations, social functioning or quality of life (∧ = 0.84, F(12, 78) = 1.19 (p > 0.05)
Disorder: depression
Group interventions
Ludman [28], USA	RCT (individual), Telephone contacts and group-based meetings (community)	Group-based chronic disease self-management programme led by a peer in addition to telephone care management provided by a counsellor. Intervention of 6 weeks with ongoing bi-monthly meetings focusing on problem solving activities	Intervention: peer-led chronic disease self-management group in addition to telephone care management + TAU (continued behavioural health care) (n = 26). Control: TAU (n = 26)	No significant differences were found among the different groups in mean SCID scores over months 6, 9 and 12 (p values not reported)
Individual interventions
Dennis [29], Canada	RCT (individual), Telephone-based/community	Individually based social support incorporating informational, appraisal (feedback) and emotional assistance until 24 weeks postpartum. Minimum of four contacts, further contact deemed as necessary	Intervention: Telephone-based peer support + TAU (standard community postpartum care from public health nurses, physicians and other providers) (n = 349) Control: TAU (n = 352)	At 12 weeks 14 % of women in the intervention group had an EPDS score >12 compared with 25 % in the control group (χ 2 = 12.5, p < 0.001). There were also significant differences in anxiety between the groups at 12 weeks (p = 0.055). No significant group differences were found in loneliness or in depression and anxiety scores at 24 weeks
Dennis [29], Canada	RCT (individual), Telephone-based/community	Individually based social support over 8 weeks incorporating informational, appraisal (feedback) and emotional assistance. Contact frequency not standardised	Intervention: telephone-based peer support + TAU (standard community postpartum care) (n = 20). Control: TAU (n = 22)	Significantly more mothers in the intervention group showed decreased depressive symptomatology at the 4 week (χ 2 = 5.18, p = 0.02) and 8 week assessment (χ 2 = 6.37, p = 0.01). Data from SES and LS not reported
Letourneau [30], Canada	RCT (individual), Home visits and telephone contacts/community	Individually based peer support over 12 weeks: Provision of informational, emotional, affirmational and practical support, weekly visits	Intervention: peer support (home visits and telephone contacts) + TAU (standard postpartum care provided by family physician) (n = 27). Control: TAU + waiting list for intervention (n = 33)	EPDS scores improved in both groups over time (F = 104, p > 0.001), but favouring the control condition (F = 5.51, p = 0.02)