Social Prescribing through Primary Care: A Systematic Review of the Evidence ()
1. Introduction
People’s health and wellbeing are determined by a range of social, psychological, economic, leisure, activity, and environmental factors (The King’s Fund, 2020) [1]. Social prescribing (SP) is a relatively recent and extensively advocated for innovation that seeks to address holistic health and wellbeing needs (Department of Health, 2006; NHS England, 2016) [2] [3]. SP has been described as “a mechanism for linking patients with non-medical sources of support within the community” (Centre Forum, 2014, p. 6) [4]. SP seeks to take a holistic view of peoples’ lives and needs by asking “what matters to you?”, and linking patients with resources and sources of support within the community (NHS England, 2021) [5].
Through connecting patients to community-based support, resources and services, SP aims to address social, health, wellbeing, functioning, leisure, activity, quality of life and economic needs, and promotes self-management (NHS England, 2016) [3]. It is viewed as a way of improving the interconnection and integration of health, social care, commercial, charity, and community resources: improving patients’ experience of services and reducing demand on primary and acute healthcare services; as well as contributing to other government objectives in relation to employment, social care, volunteering, and education/training (Wilson & Booth, 2015; The King’s Fund, 2020) [1] [6]. There is no single agreed definition of what SP entails: models of delivery, support after referral, and activities offered differ significantly across the UK (Moffatt et al., 2017) [7].
SP provides GPs with a non-medical referral option that can operate alongside existing medical treatments. GP referral is to a SP “link-worker”, often based in a GP practice; the link worker meets with a patient to define their needs, “what matters to them” and connects patients to appropriate local services or resources (Husk et al., 2016; Moffatt et al., 2017) [7] [8]. Link-worker’s roles can vary from “light touch” (referring and sign-posting), to a more holistic and instrumental approach that has a formal engagement process, sets health and wellbeing goals, facilitates behaviour change, and provides practical and emotional support over a period of time (typically three months) (Kimberlee, 2014) [9].
Randomised controlled trials (RCTs) and studies with other designs have demonstrated the positive impact of SP on patients’ health and wellbeing (Maughan et al., 2016; Bickerdike et al., 2017) [10] [11]. Qualitative evidence finds that SP services are well liked by both patients and GPs (Smith & Skivington, 2016) [12]. It is acknowledged that health and wellbeing improvements may occur over a long period of time, and that behaviour change to healthier lifestyles can often occur slowly (National Academy for Social Prescribing, 2021) [13]. The groups of people who are viewed as potentially benefiting from SP include people with mental health problems, complex needs, multiple long-term conditions, and people who are socially isolated, lack the support or financial resources they need, or who frequently attend primary or secondary health care services (Friedli et al., 2009; The King’s Fund, 2020) [1] [14]. National Institute for Health and Care Excellence (NICE) draft guidelines for mild to moderate depression recommended offering referral for exercise-related SP activities prior to prescribing anti-depressants (NICE, 2021) [15].
In 2014, a Centre Forum Mental Health Commission recommended that GP-based SP should be available in every primary care practice, to connect patients to local services and other support available in the community that can address psychosocial factors to improve patient wellbeing (Centre Forum, 2014) [4]. SP was given impetus with a commitment in the “Long Term Plan” for the NHS in England to have over 1000 trained SP link workers in post by 2020/2021 and to further expand provision so that over 900,000 people will have been referred to SP services by March 2024 (NHS, 2014) [16]. The “Long Term Plan” links SP to a wider salutogenic (generating health and wellbeing) model of delivering “universal personalised care” and adopts a more holistic approach to include a person-centred focus on wellbeing and resilience, not just the absence of disease (NHS, 2014; NHS England, 2021) [5] [16]. The plan recognises that the disease-oriented biomedical model of treatment and care cannot fully meet health and wellbeing needs, and that there needs to be a shift towards disease prevention, health promotion, self-management, and health behaviour and lifestyle modifications (Eaton et al., 2015) [17]. In September 2021, the “Accelerating Innovation in Social Prescribing” initiative was launched, seeking to enhance connections between voluntary organisations and health and social care systems, to develop SP activities that are widely accessible, have a positive impact, and reduce health inequalities (National Academy for Social Prescribing, 2021) [13].
Non-clinical community interventions such as SP should aim to result in: 1) measurable health and wellbeing benefits, and 2) cause the adoption of lifestyle behaviour change and habits (fixed action patterns [FAPs] producing automated behaviour or thoughts) that result in long-term positive outcomes; however, few interventions seem to seek to do the latter (Pretty & Barton, 2020) [18]. There is a need for all SP interventions to aim to shift personal behaviours and choices to those that enhance health and wellbeing (Pretty & Barton, 2020) [18]. To achieve behaviour change, the adoption of key components of the COM-B model of behaviour: 1) opportunity; 2) motivation: 3) capability, has been suggested (Pretty & Barton, 2020) [18].
A growing body of evidence suggests that SP can lead to a range of positive health, wellbeing, and quality of life outcomes, and reduce social isolation, depression, and anxiety (The King’s Fund, 2020) [1]. However, there remain weaknesses in the evidence base: many studies are small-scale, do not have a control group, focus on progress rather than outcomes, lack detailed descriptions of participants and interventions, have poor reporting, or relate to individual interventions rather than the SP model; and much of the evidence available is qualitative and relies on self-reported outcomes (Bickerdike et al., 2017; The King’s Fund, 2020) [1] [11]. Bickerdike et al.’s (2017) [11] review concluded that the evidence base does not allow effective assessment of who received what, for how long, with what effect, and at what cost; and there is a failure to either consider or account for potential confounding factors, undermining the ability to attribute causality (Bickerdike et al., 2017) [11]. Nevertheless, real-world evaluations have generally presented positive results (Bickerdike et al., 2017) [11].
This review is an update on the systematic review undertaken in 2016 by Bickerdike et al. (2017) [11]. Since 2016, SP has become much more widely available across the UK. This current review summarises the recent evidence for the patient-outcome effectiveness of SP programmes relevant to the UK NHS setting. In addition to review of quality, outcomes and characteristics of SP projects undertaken by Bikerdike et al., (2017) [11] we also reviewed included papers for “participant needs and referral criteria” and “Behaviour change methods” employed. This we did for both this current study’s included papers and those of Bikerdike et al., (2017) [11]. This review can inform future delivery and evaluation of SP.
2. Methods
2.1. Study Registration
The protocol is registered in PROSPERO (registration number: CRD42021265520).
2.2. Data Sources and Searches
Sources: MEDLINE, PubMed, PsycINFO (EBSCOhost), Embase, Web of Science to locate any papers describing or evaluating SP programmes. Grey literature reports of relevant evaluations in UK settings were identified via a Google search and from specific websites of organisations such as the Kings Fund, Health foundation, Nuffield Trust and NESTA. Reference lists of retrieved articles were scanned to identify additional studies. Dates for search were from 6 February 2016 (date after end of Bickerdike et al.’s [2017] [11] search) until 26 July 2021. All the searches were restricted to English language only.
2.3. Study Selection
Any published research or evaluation of programmes where healthcare professionals refer patients from a primary care setting to a SP link worker (who offers any form of SP) were eligible for inclusion. Studies were eligible if they had a comparison group or not. Search terms used: “social prescribing”, “social prescription”, “wellbeing programme”, “non-medical referral”, and “community referral”. Exclusion criteria: “literature reviews”; “studies that do not evaluate with a patient outcome measure”; “patients referred for same activities but not as part of social prescribing programme”; “no pre and post outcomes data”; and “patients referred for social prescribing activities but not from any primary care setting.”
The primary outcomes of interest were measures of mental and physical health and wellbeing, including self-reported measures. However, any other outcomes used in identified evaluations were reported. There was no restriction placed on the length of follow-up.
Papers found were deduplicated using Mendeley reference management software. Following this, papers were reviewed and separated as “social prescribing” and “non-social prescribing”, via title and abstract screening, removing non-social prescribing papers from the list. Study selection was performed by one researcher and checked by a second, with any discrepancies resolved by discussion and with recourse to a third researcher. Full text articles were then obtained and reviewed for eligibility, performed independently by two researchers, with any discrepancies resolved by discussion and with recourse to a third researcher.
2.4. Data Extraction and Quality Assessment
Data extracted were details of the setting, participants, the intervention (type, delivery mode and length of time), type of evaluation and outcomes of evaluation, participant needs, and behaviour change techniques employed. Two researchers independently undertook this, and discrepancies were resolved by discussion. Results are reported in Table 1 and Tables 3-5.
The quality assessment tool developed by the US National Heart, Lung, and Blood Institute for before-after (pre-post) was applied (National Heart, Lung, and Blood Institute, 2013) [19]. Our primary focus was on effects. This was performed independently by two researchers, with any discrepancies resolved by discussion and with recourse to a third researcher.
Results reported in Table 2.
2.5. Data Synthesis and Analysis
Given the identified study’s diversity in methodological design, outcomes, and interventions of focus, a narrative synthesis of the evidence was conducted. Data were insufficient to perform a meta-analysis for any of the outcomes. The narrative synthesis provides a description and summary of study findings and quality to investigate, and report similarities and differences between studies and patterns in the data. The results are reported along with the limitations of the methodology.
3. Results
We identified a total of 3595 records through database searches (after merging all databases). Noadditional records through other sources were found. After deduplication, 1870 titles and abstracts were screened, and 23 full-text papers were assessed for inclusion. This process resulted in eightpapers. See Figure 1: Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flow diagram.
3.1. Included Studies
We included eight evaluations of SP programmes where some form of link
![]()
![]()
![]()
Table 1. Characteristics of SP project evaluations.
worker role was used. The designs included one controlled before and after and seven uncontrolled before and after studies. In the studies, the link worker met with the patient to discuss their needs and directed them to support and resources. Details of the included evaluations are presented in Table 1.
3.2. Quality of the Evidence
Quality assessment and risk of bias for the evaluative designs are presented in Table 2. In general, evaluations did not determine sample sizes using power analysis, had significant loss to follow-up (>20%), and were lacking in completeness of outcome data. Furthermore, half had unclear selection criteria for the study population. In all cases, outcome measures were only taken once at pre and post intervention, with no follow-up. For these reasons, there was a high risk of bias.
3.3. Health and Wellbeing Outcomes
Table 3 lists outcome measures used and presents brief summaries of findings.
![]()
Table 3. Health and wellbeing outcomes.
Campaign to End Loneliness Measurement Tool (CANSAS); Camberwell Assessment of Need Short Appraisal Schedule (CANSAS); Community belonging 1-item measure from Hayward et al., 2014; EuroQol-5 Dimension (EQ-5D); Hospital Anxiety and Depression Scale (HADS); The Kessler Psychological Distress Scale (K10); Museum Wellbeing Measure for Older Adults (MWM-OA); Patient Activation Measure (PAM); Rockwood Clinical Frailty Scale (RCFS); Social support 4-item measure from Haslam et al., 2005; The Short Warwick-Edinburgh Mental Well-being Scale (SWEMWBS); UCLA 3-item Loneliness Scale, 8-item UCLA Loneliness Scale (ULS-8); The Warwick-Edinburgh Mental Well-being Scale (WEMWBS); The World Health Organisation Quality of (LifeWHOQoL).
Significant improvements were seen on some of the various measures employed. Studies reported baseline and single point follow-up outcomes only; there is no evidence based on a further follow-up point/s.
3.4. Behaviour Change Methods
Table 4 lists behaviour change methods employed and any descriptions of those
employed. Most did not state behaviour change methods employed, but of those that did goal-setting was most frequently mentioned, N = 6.
3.5. Participant Needs and Referral Criteria
Table 5 lists the needs of people seen by SP services and diagnosis and/or needs required for referral or inclusion criteria. There was great variation between projects on diagnosis and health, social and other needs inclusion and referral criteria.
4. Discussion
This systematic review examined the evidence for the patient outcome effectiveness of SP programmes relevant to the UK NHS primary care setting. The review identified eight evaluations; all of which found some evidence of positive patient outcomes. The studies reviewed show that many people who engaged in a broad range of SP activities self-reported improvements in factors such as: wellbeing, health status, quality of life, self-managing health conditions, physical activity level, and social connectedness. However, the quality of evidence was lacking due to weak methodological design (including no RCTs), high dropout, lack of post follow-up assessment of change, unclear inclusion criteria, and poor quality reporting. There was a lack of consideration and/or adjustment for potential confounding factors (e.g., other concurrent treatments or interventions), undermining the attribution of any reported positive outcomes to the SP received. This aligns with Bickerdike et al.’s (2017) [11] review conclusions and adds further evidence of generally positive results from real-world evaluations.
Bickerdike et al.’s (2017) [11] systematic review and our update found great variation in SP provision, therefore, restricting synthesis and generalisability of findings; the reviews did not establish that there is clear methodologically strong evidence that SP is or is not effective. By its nature, SP is based on the community resources identified locally, and these resources vary dependent on location (Munford et al., 2020) [41]. The underlying theories indicating that better community based social, cultural, activity, support service and green space connections
![]()
![]()
![]()
Table 5. Needs and referral criteria.
and engagement are beneficial for people are well evidenced (Pretty & Barton, 2020) [18]. In the projects reviewed, a large variety of outcome measures were employed indicating the wide potential positive impact of SP, but also a lack of focus on key overarching factors of wellbeing, self-management, and health related quality of life.
The people who were referred for and offered SP (in terms of their diagnosis and needs) varied considerably between the projects, the universal factor being that needs linked to health and wellbeing were targeted. This reflects the evidence that the underlying theories and SP interventions can be beneficial for many people in addressing their needs and factors related to specific mental and physical health diagnosis (The King’s Fund, 2020) [1]. There is an important underlying ethos of SP to address causes of poor wellbeing and health, promoting health and wellbeing and preventing illness, rather than treating illness caused (NHS England, 2021) [5]. The more universal availability of SP being applied in the NHS should allow more equitable access to SP through greater awareness, transparent referral routes, and community-based collaborations (NHS, 2014; National Academy for Social Prescribing, 2021; NHS England, 2021) [5] [13] [16].
A key component of SP is behaviour change (Pretty & Barton, 2020) [18], but we found a lack of reference to, or description and application of, behaviour change theories and practice. Only six out of 21 papers mentioned the use of goal-setting techniques, and none mentioned any other behaviour change techniques. There is strong evidence for behaviour change theories and application to improve health and wellbeing, for example “motivational interviewing” and it is important that behaviour change techniques are applied in the delivery of SP (Husk et al., 2019; Pretty & Barton, 2020) [8] [18]. Husk stated: “For all programmes, it is important to develop SP in line with complex intervention and behaviour change approaches with a careful consideration of context and capacity” (Husk et al., 2019, p. 320) [8].
5. Limitations
There were limitations of the review processes used. There is potential publication bias as other studies may exist but have not been published, and therefore, were not listed through searches employed. However, we managed to obtain a number of studies identified through requesting from the authors. There were limitations of the evidence included in the review. The intervention under review (SP) does not have a single defined inclusion or design criteria (SP intervention provision varies widely), and outcome measures vary between evaluations undertaken. These factors mean that any generalisation of findings between different SP models has to be made with caution; findings are context and intervention specific and may not be transferable to other SP models.
6. Conclusions
Further expansion of primary care-based SP requires a strong evidence base, to define what works, for who, and how (Munford et al., 2020) [41]. Further research should seek to gain a deeper understanding of the application, complexity, challenges, and successes of the SP model and delivery through a review of qualitative evidence and interviewing SP link workers and patients referred through primary care. There is a need to improve the ways by which SP schemes are evaluated, perhaps through systematic evaluation of SP in a particular NHS region; funding and NHS provider support would be required to do this. This could enable SP to be improved and be more cost-effective.
Bickerdike et al. (2017) [11] suggested the design and adoption of a common evaluation and analytical framework (e.g., Lamont et al., 2016) [42] and systematic reporting (e.g. Standards for Quality Improvement Reporting Excellence [SQUIRE] [Ogrinc et al., 2016] [43] ) to enable standardised metrics, cross-site comparison, transparency, and shared learning. As in NHS Improving Access to Psychological Therapies (IAPT) services, a “Minimum Data Set” (MDS) could be adopted with measures that would include, for example, the WEMWBS, PAM, and EQ-5D-5L, to assess three key factors of wellbeing, self-management, and health-related quality of life. It is vital that SP services seek to collaboratively work with patients to effectively define and assess outcomes and apply behaviour change techniques to enable healthier lifestyles in the short and long term.
What Is Known about This Topic?
1) Social prescription is based on evidenced theories indicating that social, psychological, economic, leisure, activity, and environmental factors impact on health and wellbeing;
2) Methologically strong research and evaluation is lacking in social prescription;
3) Social prescription provision is rapidly expanding in the UK’s NHS and is being applied internationally.
What This Paper Adds?
1) There is evidence of positive impact on outcomes through patient self-reported health, wellbeing, self-management, sociability, and functioning measures;
2) There is a need to improve evaluation through the adoption of common outcome measures and evaluation/analytical frameworks;
3) It is vital that social prescription applies behaviour change techniques to enable healthier lifestyles in the short and long term.
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.
Funding Information
This report is independent research funded by the National Institute for Health Research (NIHR) Clinical Research Network (CRN).