Background Rates of readmission are high following discharge from psychiatric inpatient care. Evidence suggests that transitional interventions incorporating peer support might improve outcomes. Peer support is rapidly being introduced into mental health services, typically delivered by peer workers (people with experiences of mental health problems trained to support others with similar problems). Evidence for the effectiveness of peer support remains equivocal, and the quality of randomised controlled trials to date is often poor. There is an absence of formal cost-effectiveness studies of peer support in mental health services. Objectives This programme aimed to develop, pilot and trial a peer support intervention to reduce readmission to inpatient psychiatric care in the year post-discharge. The programme also developed a peer support fidelity index and evaluated the impact of peer support on peer workers. Design Linked work packages comprised: (1) systematic review and stakeholder consensus work to develop a peer support for discharge intervention; (2) development and psychometric testing of a peer support fidelity index; (3) pilot trial; (4) individually randomised controlled trial of the intervention, including mixed methods process evaluation and economic evaluation; (5) mixed method cohort study to evaluate the impact of peer support on peer workers. The research team included: two experienced service user researchers who oversaw patient and public involvement; service user researchers employed to develop and undertake data collection and analysis; a Lived Experience Advisory Group that informed all stages of the research. Setting The programme took place in inpatient and community mental health services in seven mental health National Health Service trusts in England. Participants Participants included 590 psychiatric inpatients who had had at least one previous admission in the preceding 2 years; 32 peer workers who delivered the intervention; and 8 peer workers’ supervisors. Intervention Participants randomised to peer support were offered at least one session of manualised peer support for discharge prior to discharge and then approximately weekly for 4 months post-discharge. Main outcome measures The primary outcome for the trial was readmission (formal or informal) to psychiatric inpatient care (readmitted or not) within 1 year of discharge from the index admission. Secondary outcomes included inpatient and emergency mental health service use at 1 year post discharge, plus standardised measures of psychiatric symptom severity and psychosocial outcomes, measured at end of intervention (4 months post discharge). Data sources Service use data were collected from electronic patient records, standardised measures of outcome and qualitative data were collected by interview. Review methods We produced two systematic reviews of one-to-one peer support for adults in mental health services. The first included studies of all designs and identified components of peer support interventions; the second was restricted to randomised controlled trials and pooled data from multiple studies to conduct meta-analyses of the effects of peer support. Results Our systematic review indicated that one-to-one peer support improved individual recovery and empowerment but did not reduce hospitalisation. The main trial demonstrated that one-to-one peer support did not have a significant effect on readmission. There was no significant reduction in secondary service use outcomes at 1-year, or improvement in clinical or psychosocial outcomes at 4 months. Participants who received a pre-defined minimal amount of peer support were less likely to be readmitted than patients in the control group who might also have received the minimal amount if offered. Compared to care as usual, black participants in the intervention group were significantly less likely to be readmitted than patients of any other ethnicity (odds ratio 0.40, 95% confidence interval 0.17 to 0.94; p = 0.0305). The economic evaluation indicated a likelihood that peer support offered a reduction in cost in excess of £2500 per participant compared to care as usual (95% confidence interval −£21,546 to £3845). The process evaluation indicated that length and quality of first session of peer support predicted ongoing engagement, and that peer support offered a unique relationship that enables social connection. The impact study indicated that peer workers found their work rewarding and offering opportunities for personal growth but could find the work emotionally and practically challenging while expressing a need for ongoing training and career development. Limitations In the trial, follow-up rates at 4 months were poor, reducing confidence in some of our analyses of secondary outcome and in a wider societal perspective on our health economic evaluation. Conclusions One-to-one peer support for discharge from inpatient psychiatric care, offered in addition to care as usual to participants at risk of readmission, is not superior to care as usual alone in the 12 months post-discharge. Future work Further research is needed to optimise engagement with peer support and better understand experiences and outcomes for people from black and other ethnic communities. Study registration The systematic review is registered as PROSPERO CRD42015025621. The trial is registered with the ISRCTN clinical trial register, number ISRCTN 10043328. Funding This award was funded by the National Institute for Health and Care Research (NIHR) Programme Grants for Applied Research programme (NIHR award ref: RP-PG-1212-20019) and is published in full in Programme Grants for Applied Research; Vol. 11, No. 8. See the NIHR Funding and Awards website for further award information. Plain language summary When people share similar experiences of mental health problems and provide each other with emotional or practical support this can be called peer support. In recent years peer support has been introduced into mental health services. The months after discharge from psychiatric hospital can be difficult for some people. There is research that suggests that peer support might be helpful for people at discharge. There is a lack of good quality trials of peer support in mental health services. There are no studies that tell us if peer support provides value for money. We developed new training for peer workers and a handbook guiding peer support for discharge. We tried this out in two mental health National Health Service trusts to check that we could deliver the peer support and collect the information we needed for our trial. We recruited 590 people from seven mental health National Health Service trusts while they were staying on psychiatric wards. Half of those people were randomly selected to receive peer support for discharge, and half to receive community mental health care only. We found that people offered peer support were just as likely to be readmitted to hospital in the year after they were discharged as people offered usual care only. We found that people who met their peer worker at least twice were less likely to be readmitted in the next year. Black people offered peer support were much less likely to be readmitted compared to people offered usual care than other people in the study (although numbers were small). Our results also suggested that cost of care could be slightly less for people receiving peer support. We interviewed some people who had been offered peer support who told us what they valued about peer support. We interviewed peer workers who told us the work was rewarding but could be challenging. Scientific summary Background Risk of self-harm, suicide and readmission is high in the months immediately following discharge from psychiatric inpatient care. There is some evidence that transitional interventions incorporating peer support improve outcomes, including reduced readmission rates. Peer support is rapidly being introduced into mental health services internationally, typically delivered by peer workers (PWs) (people with their own experiences of mental health problems trained to support others with similar problems). Evidence for the effectiveness of peer support in mental health services remains equivocal, and the quality of randomised controlled trials to date is often poor, with trials at risk of bias because of unclear randomisation and masking procedures and incomplete reporting of outcomes. Peer support interventions are often poorly described, including the support provided for PWs, and the mechanisms of peer support unclear. Health economic evaluations are absent from the literature. Objectives The aim of the programme is to manualise, pilot and trial a peer support intervention to enhance discharge from inpatient to community mental health care, significantly reducing readmissions and the associated cost of care. The detailed research objectives of the programme are: to refine an empirically and theoretically grounded model that explains how peer support impacts on outcomes for service users post-discharge to develop and manualise a peer support intervention to enhance discharge to develop an index to assess the fidelity of peer support interventions to conduct a high-quality randomised controlled trial of the intervention to establish the effectiveness and cost-effectiveness of a peer support intervention to enhance discharge to explore the impact on PWs of working in a peer support role. Methods A training manual for PWs and intervention handbook were developed in work package 1 (months 1–15). We produced two systematic reviews of one-to-one peer support for adults in mental health services. The first, until April 2015, included studies of all designs and was used to identify components of peer support interventions; the second, until June 2019, was restricted to randomised controlled trials and pooled data from multiple studies to conduct meta-analyses of the effects of peer support. In both searches we searched MEDLINE, PsycINFO, Embase, CINAHL and Cochrane databases (as well as grey literature for the review of components), using ‘peer’, ‘consumer’, ‘survivor’ or ‘prosumer’ adjacent to ‘support’, ‘supporter’, ‘provider’, ‘worker’, ‘specialist’, ‘consultant’, ‘tutor’, ‘educator’, ‘mentor’, ‘intervention’, ‘listener’, ‘mediator’, ‘counsellor’, ‘befriender’ or ‘therapist’ as search terms for peer support. Paper selection at both title and abstract, and full text stage was undertaken independently by two researchers, with data extracted to spreadsheets relevant to each review. Results for the first review were combined with consensus workshops with stakeholder panels to develop and refine the peer support for discharge intervention. We also developed and tested the appropriateness, inter-rated reliability and internal consistency of a peer support fidelity index (work package 2, months 7–24). The index was first tested in 20 peer support projects not involved in the trial, before being retested and then implemented at three timepoints in each trial site. We conducted an internal pilot of the intervention and trial procedures in two sites (work package 3, year 2 of the programme) before proceeding to full trial in seven sites (work package 4, years 3–5). Sites were inpatient and community mental health services in seven mental health National Health Service trusts in England. Participants were people admitted to psychiatric inpatient care who had at least one previous admission in the preceding 2 years (i.e. who were at increased risk of readmission), the PWs who were employed and trained at each site to deliver the intervention, and the PW co-ordinators (PWCs) who supervised PWs at each site. Participants randomised to peer support were offered a manualised peer support for discharge intervention and care as usual (CAU). Participants in the control group received CAU only (discharge summary sent to community or primary care mental health team who contacted participant within 1 week of discharge to plan ongoing care). The primary outcome for the trial was readmission (formal or informal) to psychiatric inpatient care (readmitted or not) within 1 year of discharge from the index admission. Secondary outcomes were number of voluntary admissions, involuntary admissions and total number of admissions, total number of days in hospital, time to first readmission, use of accident and emergency services for a psychiatric emergency (measured as number of episodes of liaison psychiatry contact) and number of contacts with crisis resolution and home treatment teams in the year post discharge, plus standardised measures of psychiatric symptom levels, subjective quality of life, social inclusion, hope for the future and strength of social network, measured at end of intervention (4 months post discharge). Analyses were conducted on an intention-to-treat (ITT) basis. We carried out a Complier Average Casual Effect (CACE) analysis of the primary outcome (where compliers were participants who had at least two PW meetings, at least one of which was in the community following discharge). The CACE was estimated with a two-stage estimation procedure. In the first stage, a logistic regression of treatment receipt regressed on randomisation was conducted. In the second stage, a Poisson regression of the outcome on treatment receipt was conducted. The analysis was adjusted for the same covariates as the ITT analysis. A bootstrap (1000 samples) was used to obtain bias corrected and accelerated confidence intervals (CIs). Subgroup analyses for the primary outcome were pre-specified: ethnicity (any black ethnicity, all other ethnicities); primary diagnosis at index admission (psychotic disorders, personality disorders, other eligible disorders); first language (English, other). Service use data were collected from electronic patient records (EPRs) at study site, standardised measures of outcome and qualitative data were collected by face-to-face interview by a member of the study team. Our primary economic analysis of total costs over 12 months (using EPR of mental health service use), allowing for cost of peer support, and the secondary cost-effectiveness analysis at 4 months [using self-reported quality of life to derive quality-adjusted life-years (QALYs)] were carried out from an NHS mental health service perspective. A wider ‘societal’ perspective was taken when analysing non-NHS mental health care costs over 4 months (using self-reported service use outside of mental health NHS care). All analyses were conducted on an ITT basis using generalised linear modelling (GLM) with a logarithmic link function. We conducted a mixed methods process evaluation as part of work package 4, alongside the trial. Quantitative analyses used regression models to explore pre- and post-randomisation predictors of engagement with the peer support intervention. In-depth qualitative interviews with a subsample of 39 trial participants and all 32 PWs who were involved in delivering the intervention explored their experiences of peer support and sought to elucidate and refine the change model underpinning the intervention. Interviews were undertaken by service user researchers and data analysed using a ‘co-production’ approach to integrate the full range of perspectives on the research team – clinical, academic and experiential – in interpretive workshops, producing an analytical framework that was then used to code the full set of qualitative interviews. Work package 5 was a mixed method, longitudinal cohort study which explored the impact of providing peer support on PWs (from month 31 until the end of year 5). PWs completed standardised measures of well-being and employment outcomes, and in-depth interviews at three timepoints. All data were collected by service user researchers, who played a key role in developing interview schedules, refining rating procedures for the fidelity index, and analysis and interpretation of qualitative interview data. Development of the intervention and all research procedures were informed by a Lived Experience Advisory Panel (LEAP), as well as the experiential knowledge brought by service user researchers and PWs on the research team. Results In the first (2015) systematic review we identified 97 studies (including 12 from grey literature) that contributed 44 potential components to the intervention development process. A total of 66 potential components were identified when combined with outputs of our stakeholder workshops. Through iterative rounds of consensus building and testing we developed and refined our peer support intervention, comprising a bespoke PW training programme and a detailed handbook guiding implementation. In the second (2019) review we identified 23 studies reporting 19 trials. We found that one-to-one peer support in mental health services has a small but statistically significant benefit for individual recovery [standardised mean difference (SMD) 0.22, 95% CI 0.01 to 0.42; p = 0.042] and empowerment (SMD 0.23, 95% CI 0.04 to 0.42; p = 0.020). There was no effect on clinical outcomes such as symptoms or hospitalisation; the risk of being hospitalised was reduced by 14% for those receiving peer support but was non-significant [risk ratio (RR) 0.86, 95% CI 0.66 to 1.13]. We developed a principles-based fidelity index that had good acceptability and psychometric properties. The index measures set-up, delivery and overall fidelity of peer support against four domains (principles): building trusting relationships based on shared lived experience; reciprocity and mutuality; leadership, choice and control; building strengths and making connections to community. In the trial, fidelity was good at set-up in all sites, while fidelity of delivery of peer support was lower in sites where PWs were employed in voluntary sector organisations outside of the NHS. The pilot trial indicated that progression criteria were met subject to a small number of actions to improve the rate of recruitment of participants. In the main trial we successfully recruited our target of 590 participants, and participant characteristics were well-balanced between groups. In the PW group, 136 (47.4%) participants were readmitted to psychiatric inpatient care within 12 months post-index admission, and 146 (50.2%) in the CAU group. The adjusted relative risk of readmission in the ITT analysis was 0.97 (95% CI 0.82 to 1.14; p = 0.6777), and the adjusted odds ratio (OR) was 0.93 (95% CI 0.66 to 1.30). In the CACE analysis, the relative risk of readmission according to the natural indirect effect (RR 0.88, 95% CI 0.76 to 0.99) was lower than from the ITT analysis and was significant. In subgroup analyses (see Table 6), for patients of any black ethnicity the adjusted OR of readmission was 0.40 (95% CI 0.17 to 0.94), while for any other ethnicity the OR was 1.12 (95% CI 0.77 to 1.63; interaction p = 0.0305). There were no statistically significant differences between the groups in any of the secondary outcomes assessed at 4 or 12 months. Adherence to the intervention was assessable in 268 (91.2%) participants with a mean of 1.8 [standard deviation (SD) = 2.9] face-to-face contacts with a PW in hospital, 4.4 (SD = 4.6) post discharge. There was a total of 67 serious adverse events (SAE) reported in the trial (34 in the peer support group, 33 in the CAU group) from 51 participants (26 in the peer support group, 25 in the CAU group). One SAE in the peer support group, an incident of self-harm, was reported as related to the intervention. Number and type of SAE included 12 deaths, none of which were reported as related to the study. A cost analysis of mental health service contacts over a 12-month period following discharge from inpatient care showed that, adjusting for baseline covariates, exposure to peer support was associated with a reduction in mean total costs of £2631 (95% CI −£21,546 to £3845): this amounted to a 10% reduction in mean total costs over 12-month post hospital discharge compared to usual care (95% CI −31% to 15%). Given sampling uncertainty there was an estimated 82% probability that peer support was associated with lower total costs over 12 months (or a 18% chance that usual care was the lower cost alternative). Most of the cost advantage over follow-up was due to reductions in the cost of bed day utilisation. Over 4 months, and considering patient quality of life outcomes as well as cost, peer support was also found to be cost-effective from an NHS mental health service perspective. The expected QALY gains associated with peer support were marginal: a 0.002 QALY improvement per participant, equivalent to less than a single day in full health. Participants who were heterosexual were less likely to engage with peer support than gay, lesbian or bisexual participants, OR 0.3 (95% CI 0.08 to 0.87; p = 0.029). We found that length of first contact (in minutes) was positively associated with engaging with peer support, OR 1.02 (95% CI 1.00 to 1.04; p = 0.010), and participants who went on to engage with peer support experienced more relationship building activity in that first contact. A shorter period between allocation to peer support and discharge (in days) was also associated with engagement in the intervention, OR 0.99 (95% CI 0.98 to 1.00; p = 0.002). Our qualitative process evaluation largely supports our original change model, further elucidating the distinctive nature of the peer-to-peer relationship while indicating the role that PWs played in enabling people to build relationships and make connection to community. Choice and control over how people engaged with peer support was identified as an important mediator of good experience and outcomes of peer support. Peer worker well-being, job satisfaction, team working and burn out scores were close to or better than scores for appropriate norm populations throughout the study. There were small but significant drops in well-being, personal satisfaction and satisfaction with workload after 4 months in post, and a similar small increase in burn out, but these changes were not maintained at 12 months. Satisfaction with training and job prospects were lower at 12 months. Qualitative data largely confirmed these findings with PWs indicating that they found that peer support work could be emotionally and practically challenging for the first few months but was rewarding and offered opportunities for personal growth. Conclusions We conclude that peer support for discharge, offered to participants at risk of readmission, was not superior to CAU. Peer support should not be commissioned with the expectation that it reduces readmissions for this group, although some cost-saving based on fewer days in hospital is likely. Our trial findings reflect those of our systematic review, suggesting that one-to-one peer support in mental health services is unlikely to improve clinical outcomes or reduce hospitalisation. Further research is needed to: improve implementation in order to optimise engagement with peer support establish the impact of peer support on psychosocial outcomes understand and evaluate the impacts of peer support for people from different ethnic communities (and especially to optimise benefits for black people using mental health services) understand and evaluate the impacts of peer support for people from different clinical populations establish the construct validity of our fidelity index through use with larger samples better understand the ongoing training and support needs, and career development pathway for PWs. Study registration The systematic review is registered as PROSPERO CRD42015025621. The trial is registered with the ISRCTN clinical trial register, number ISRCTN 10043328. Funding This award was funded by the UK National Institute for Health and Care Research (NIHR) Programme Grants for Applied Research programme (NIHR award ref: RP-PG-1212-20019) and is published in full in Programme Grants for Applied Research; Vol. 11, No. 8. See the NIHR Funding and Awards website for further award information. This paper presents independent research funded by NIHR. The views expressed are those of the authors and not necessarily those of the UK National Health Service (NHS), the NIHR or the Department of Health and Social Care.