by Marija Trachtenberg and Michael Parsonage

23 July 2013

Peer support is the provision of practical and emotional help for people with long-term mental health problems by people with their own lived experience of such problems. The underlying idea is not a new one. For example, there is evidence going back to the 1840s of peer support in the old asylums, provided by Alleged Lunatic Friends (Hervey, 1986). Indeed, for as long as people have used mental health services, they have provided each other with friendship, mutual support and shared coping strategies, but it is only relatively recently that the value of such relationships has been formally recognised in the development of paid employment roles. Peer support workers may be employed in place of traditional staff, for example as case managers, or – more commonly – as additional members of a mental health team, fulfilling such roles as befriending, mentoring, social support and advocacy.

Increasing numbers of peer support workers are now being employed in mental health services, both in this country and elsewhere, but good quality evidence on the effectiveness of this form of service delivery is in short supply and even less is known about its cost-effectiveness. A recent Cochrane review was able to identify only 11 randomised controlled trials and came to largely agnostic conclusions (Pitt et al. 2013). Other reviews have taken a more inclusive approach, including evidence not derived from RCTs, and a number of broadly positive conclusions on effectiveness have started to emerge (Repper and Carter, 2011). These include benefits not just for the service users who receive peer support but also for the peer workers themselves and at the organisational level, where the introduction of peer support is increasingly seen as a powerful way of promoting a more recovery-focused approach to mental health care. In no study has the employment of peer support workers been found to result in worse health outcomes for those receiving the service and, although not replicated in all studies, findings have been reported of improvements across a range of outcomes including: patients’ feelings of empowerment and self-confidence; self-reported physical and emotional health and clinician-assessed global functioning; satisfaction with services; community integration and social functioning; stability in employment, education and training; and reduced alcohol and drug use among patients with co-occurring substance abuse problems.

A first attempt at assessing whether peer support provides value for money is set out in a paper just published by the Centre for Mental Health (Trachtenberg et al. 2013) and launched at this year’s annual conference of the NHS Confederation. The analysis looks specifically at whether peer support workers can reduce psychiatric inpatient bed use, either by preventing admissions or by reducing length of stay. We chose to look at the impact on inpatient bed use partly for reasons of data availability but also because of the very high unit cost of hospital care and because this has been an area where service users’ experience of care has consistently been reported to be rather poor (Care Quality Commission 2009) and thus it seemed an area where the impact of peer workers might be particularly strong. If peer support can improve patients’ feelings of empowerment, self-esteem and confidence, this may help them to manage their lives in the community better, with a correspondingly reduced need for inpatient care.

We identified six studies in the research literature which give some evidence on the relationship between peer support and inpatient bed use. Five of these were from the US and one from Australia. Only two were RCTs, although three of the others did have some form of comparison group.  Study authors were contacted in all cases for additional information and on this basis we were able to produce estimates of the number of hospital bed-days saved per full-time equivalent peer support worker in each study. These figures on bed-days saved and peer worker input were then converted to £s using unit costs for England in 2011-12, resulting in a simple benefit:cost ratio of the following form: value of hospital bed-days saved per peer support worker divided by cost per support worker.

Four of the six studies show a benefit:cost ratio substantially in excess of one, i.e. the value of bed-days saved comfortably exceeds the cost of employing peer workers. In one study, the ratio is positive but less than one (some savings but not enough to fully offset the costs of employment) and in one the ratio is negative, implying that in this case the use of peer workers is associated with an increase rather than a reduction in bed use. Aggregated across all six studies, the benefit:cost ratio works out at 4.8:1 on a weighted average basis (with weights determined by the size of each study’s patient sample). The overall conclusion suggested by these figures is that peer workers bring about significant reductions in hospital bed-use among the patients they support, leading to savings which are well in excess of additional pay costs. These savings are of course over and above any positive impact on non-financial outcomes relating to the mental health and quality of life of service users.

A number of major limitations must be acknowledged. First, the evidence base is extremely modest, both in scale and in quality. Second, there is a good deal of variation between the studies in the nature of the intervention being evaluated (e.g. the roles played by peer workers, the amount of training they received and the frequency and duration of their contacts with patients). Third, no allowance is made for the possible impact of peer support on service costs other than hospital bed use. And fourth, none of the studies is from this country.

Notwithstanding these and other shortcomings in the analysis, our conclusion is that, overall, the available evidence is sufficient to justify continuing developments in the use of properly trained and supported peer workers in mental health teams, alongside more research evaluating their effects.     


Hervey N (1986) Advocacy or folly: the Alleged Lunatic Friends SocietyMedical History, 30, 245-275.

Pitt V, Lowe D, Hill S et al. (2013) Consumer providers of care for adult clients of statutory mental health services, Cochrane Database of Systematic Reviews 2013, issue 3.

Repper J, Carter T (2011) A review of the literature on peer support in mental health servicesJournal of Mental Health, 20, 4, 392-411.

Trachtenberg M, Parsonage M, Shepherd G, Boardman J (2013) Peer support in mental health care: is it good value for money?Centre for Mental Health, London.

Care Quality Commission (2009) Mental health acute inpatient services survey 2009,Care Quality Commission, London. 


Marija Trachtenberg is a graduate of the LSE Masters course in health economics and has worked as a research assistant at the Centre for Mental Health.

Michael Parsonage is Chief Economist, Centre for Mental Health.


This post was first published on the LSE Health and Social Care blog (July 15 2013)