Introduction

The prevalence of complex health and social needs are increasing demands on Australian primary healthcare services [1] and General Practice/practitioners (GPs) are rapidly reaching capacity [2]. There are also notable disparities in healthcare service access, particularly among vulnerable population groups [2–5]. Existing healthcare systems lack the capacity to comprehensively address the social determinants of health, including social participation, employment, education, neighbourhood conditions, and other social and economic factors [6], which influence health and well-being [7]. Thus, additional approaches are imperative to advance person-centred care and improve health outcomes [8].

Social prescribing offers a person-centred and scalable solution to unmet health and social needs, by leveraging community assets to facilitate collaboration across health, social and community services [8, 9]. This multidisciplinary approach addresses healthcare system constraints without depending solely on routine medical care. UK evaluations have consistently found positive social return on investment for social prescribing schemes [10, 11], with a similar outcome found in Australia [12]. However, these evaluations primarily assess costs to the health system and overlook expenses related to non-medical service utilisation. Ultimately, the person-centred and collaborative nature of social prescribing empowers individuals by equipping them with the necessary knowledge, skills, motivation, and confidence to manage their own health and social well-being, promoting sustained benefits [8].

Globally, social prescribing is implemented through two main models: the direct referral model and the link worker model. In the direct referral model, health professionals (e.g. GP, community worker, social worker) refer patients directly to non-medical community-based programs or activities to improve health and well-being [9]. The link worker model involves a referral to a link worker, who assists the person in developing and executing a personalised plan that utilises community-based resources to enhance their quality of life [13]. This model also focuses on building the individual’s confidence, skills and health literacy to manage their own unmet health and social needs, such as chronic conditions and loneliness [14]. Both models facilitate community engagement and reduce loneliness by promoting engagement in physical activity, arts, culture and the natural world, or promoting access to practical advice and information [15].

In Australia, both models of social prescribing operate, however the link worker model presents several advantages [16]. Firstly, link workers possess a higher level of knowledge of local services compared to GPs, and can dedicate additional time to further engage with clients and assess which services, programs, and activities would be most appropriate [14]. Secondly, link workers can oversee follow-up on health-related non-medical challenges without the cost of an additional GP appointment, with the potential for integrated GP feedback systems [17, 18]. Given Australia’s diverse demographics, where 27.6% of the population is born overseas [19] and 33.1% reside in regional or rural areas [20], the link worker model is vital. This model is better equipped to accommodate the diverse social determinants of health inherent in the population characteristics and address the wide spectrum of health and social care needs [8, 9].

Social prescribing has gained traction globally over the past decade [8], with a recent review identifying 159 social prescribing programs for adults internationally [21]. In Australia, there is no funding model or national framework, however the adoption of social prescribing is rapidly increasing [15]. This trend aligns with the Australian government’s ‘National Preventive Health Strategy 2021 to 2030’, which advocates for the integration of social prescribing to prevent public health concerns, and Australia’s 10 Year Primary Health Plan 2022–2032, which charges Primary Health Networks to trial models of social prescribing nationally. Currently, the Mitchell Institute is conducting a feasibility study commissioned by the Australian Department of Health and Aged Care to explore social prescribing within primary health care. While international experiences provide valuable insights, local factors, such as the remoteness of the Australian outback and regionality, culturally and linguistically diverse groups and differing funding systems across states and territories, necessitate a tailored approach to implementation within the Australian context.

Although international guidance exists for planning and implementing social prescribing programs [21], evaluating these programs remains a consistent challenge [15]. This is because social prescribing programs consist of many components and are complex, often consisting of individualised interventions that can range in type, purpose, number, duration and environment of delivery [1, 8]. There is a need for a ‘multi-criteria’ approach to effectively assess both the implementation process and the outcomes of social prescribing interventions [1]. However, challenges to this approach include data inconsistency, diverse measurement with limited comparability, difficulty in identifying influential aspects of the intervention, and inflexibility in the selection of evaluation strategies, all of which compromise data quality [1, 9, 22].

Consequently, to enhance the evaluation of social prescribing programs in the Australian context, a crucial initial step involves establishing standardized data collection protocols that are endorsed by stakeholders [22, 23]. This necessitates consensus on the outcome domains to be investigated and a directive for the adoption of validated outcome measures [23]. In this integrative review, we defined outcome domains as “broad, descriptive categories under which several, more specific outcomes might be grouped” [24] and outcome measures as “the tool used to assess the outcome” [25]. Establishing core outcome domains for key diagnostic groups and interventions is paramount as it allows for the aggregation and comparison of findings across studies, facilitating evidence-informed decision-making [26]. A core outcome set defines a standardised set of outcomes to be routinely measured and reported in intervention studies [24, 27, 28]. Despite the importance of core outcome sets, one has yet to be defined for social prescribing. Therefore, identifying and examining outcome domains and measures is an initial step to enable robust evaluation of social prescribing programs in Australia.

Aims and objectives

This research aims to systematically explore the Australian social prescribing literature, focusing on the link worker model. We seek to chronicle the current Australian research, which will serve as a foundational reference to track social prescribing progression. By detailing ‘what’ has been measured and ’ how’, we aim to inform future research, guide evaluation practices, and influence policy decisions in the early stages of scaling up link worker social prescribing in Australia. Specifically, our objectives are to identify and describe:

1. outcome domains and outcome measures used to evaluate social prescribing programs in Australia;
2. methodological approaches and evaluation designs, lending insights into the foundational research methodologies employed.

Method

An integrative review was conducted according to the methodology of Whittemore and Knafl [29] and includes both qualitative and quantitative studies in order to provide a comprehensive and nuanced understanding of current evidence. Studies were identified using academic and clinical trial database searches and reference list checking, and evaluated using the Mixed Methods Appraisal Tool [MMAT; 30].

Selection of sources of evidence

All database searches were conducted by author TT with results imported into Endnote X9 [31]. Inclusion and exclusion criteria are presented in Table 1. Initial double screening of identified records according to selection criteria was completed based on record titles and abstracts (conducted independently by authors TT, CT) and disagreement resolved by discussion. Full-text review of the remaining papers was conducted independently by four authors (CA, JB, TT, CT) and disagreement was resolved by discussion. Protocol papers were included except when the corresponding article with results could be accessed. In this scenario, only the article with results was included. Social prescribing link workers were defined as people who liaise between clients, health professionals and community organisations to connect people to community-based support, including activities and services that meet practical, social and emotional needs [13].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Inclusion and exclusion criteria.

https://doi.org/10.1371/journal.pone.0309783.t001

Data extraction

A data extraction tool was developed in Microsoft Word [32] based on the template by Silver and Francis [33]. Study details were systematically extracted, summarised, and displayed to facilitate comparison and synthesis [29]. Data items were: Citation, Aim, Design, Setting, Sample, Intervention, Comparator, Domains, Outcome Measures, and Findings (Table 2). Data was extracted independently by two authors (TT, CT) and inconsistencies were resolved through team discussion. For qualitative studies, an additional two authors (CA, JB) also reviewed extracted Domains and Outcome Measures to ensure accuracy of interpretation.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Characteristics of included studies.

https://doi.org/10.1371/journal.pone.0309783.t002

Results

As shown in the PRISMA diagram (Fig 1), the selection procedure initially identified 177 records from academic databases and twenty-seven records were removed, leaving 150 unique records for screening. Title and abstract screening excluded 136 records as they did not describe social prescribing interventions (n = 92), were not set in Australia (n = 17), or did not describe primary research, including reviews (n = 14), and other article types (n = 13), for example, commentaries and clinical management guidelines. Seventeen review papers’ reference lists were scanned including [1, 15, 17, 21, 22, 34–45]. Fourteen full-text reports were retrieved (all were available) and assessed for eligibility. Seven reports were excluded as they were not social prescribing (n = 5), did not include the link worker model of social prescribing (n = 1), was a case study (n = 1), and were not set in Australia (n = 1). Six studies were eligible, including three published studies [46–48], and three study protocols [49–51]. A summary of the papers comprised in this integrative review and their corresponding MMAT score are displayed in Table 2.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Study selection PRISMA diagram.

From: Page, McKenzie [52].

https://doi.org/10.1371/journal.pone.0309783.g001

Outcome domains and measures

Ten outcome domains were identified among the included studies and these were categorised as ‘person-level’ (n = 6), ‘system-level’ (n = 2), and ‘program implementation’ domains (n = 2) (see Table 3). Two studies [46, 48] included all three domains, one study [47] and two study protocols [49, 50] included ‘person-level’ and ‘system-level’ domains, and one study protocol [51] included ‘person-level’ and ‘program implementation’ domains. The primary outcome across all studies and protocols was the person-level domain, in which global well-being, social and emotional well-being were most commonly evaluated, with the WHO Quality of Life Brief Assessment and the short form UCLA Loneliness Scale the most commonly applied standardised outcome measures. At the system level, health service utilisation was mainly evaluated (see Table 3).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Summary of outcome domains and measurements.

https://doi.org/10.1371/journal.pone.0309783.t003

A mixed methods study by Aggar, Caruana [46] exploring social prescribing for people with work-related injuries and psychosocial difficulties was the only study that reported on quality and interactions of the link workers. This integrative review revealed that positive relationships between participants and their link worker improved outcome measures. Woolfenden [51] is the only study that proposes to evaluate link workers through surveys and qualitative methods as part of their study design.

Discussion

This integrative review identified only six published link worker social prescribing Australian studies, of which three were study protocols. The studies varied in their methodological approaches and study designs. Across these six papers, ten outcome domains were generated and categorised as: ‘person-level’, ‘system-level’, and ‘program implementation’ domains. Across these three outcome domains, a total of 46 outcome measures were applied. The variation in outcome domains and measures makes sense in light of the individualised approach of social prescribing but presents a challenge in terms of a core minimum dataset for the purposes of benchmarking across social prescribing programs. Overall, our findings illustrate the perceived potential for social prescribing to positively impact a range of participant groups, but there is inconsistency regarding both ‘what’ to measure and ‘how’ to measure it. This integrative review reports on the current state of evidence in Australia, with the potential to track changes and trends over time.

Despite the variation in participant groups, the primary aim in all studies was the ‘person-level’ domain, in which global well-being and social well-being were consistently evaluated. At the ‘system level’, health service utilisation was most commonly evaluated. These findings reflect international literature [44], including the recommendation for the evaluation of ‘person level’ and ‘system level’ outcome domains for social prescribing programs in the UK [41]. Future Australian studies should adopt the dual focus of ‘person level’ and ‘system level’ outcome domains to further develop the evidence base for social health provision in the context of an already stretched health system.

While global well-being and social well-being were evaluated across all studies, a variation of outcome measures were applied, the most common were the WHO Quality of Life Brief Assessment and the short form UCLA Loneliness Scale. The measurement of social well-being in this review was highly heterogeneous, a common construct and measurement challenge within the field of social connections [39]. The nine social measures identified in this integrative review provide nuanced insight into the broad range of social prescribing programs that aim to address a range of social dimensions, as one aspect of the broader social determinants of health. It is worth noting that the social determinants of health and equity are often identified as key policy drivers for funding social prescribing 36], however specific measurement of the broader social determinants of health (e.g., utilising tools like The Protocol for Responding to and Assessing Patients’ Assets, Risks, and Experiences [PRAPARE]) was under-represented in the Australian literature. While diverse outcome measures demonstrate a commitment to addressing the varied healthcare needs of different population groups and individuals, heterogeneity poses major challenges for the synthesis, comparison, and evaluation of social prescribing programs [23]. Particularly, standardised outcome measures will be required for meta-analyses, which comprehensively synthesise current evidence and are used for evidence-based decision-making in healthcare and policy [23]. Additionally, this study’s findings suggest that the current Australian evidence base is limited, which highlights the need for further research and critical assessment of the findings. Consequently, there is a need for uniformity and the use of validated measures for the core or common aims of social prescribing (e.g. quality of life) to which most programs aspire. This is essential to benchmarking and facilitating well-informed decision-making, allowing for the optimal expansion of the link worker model of social prescribing within the Australian health landscape. At the same time, there is a need for flexibility to use more specific tools that align with the more specific goals of individual programs (e.g., return to work) and inclusion of the participant’s expectation of participating in a social prescribing program.

Despite the exponential growth of research in other countries, this integrative review has identified that social prescribing in Australia is in its infancy, with only a small number of studies addressing the needs of a heterogeneous group of clients. Given the limited rollout of the link worker model of social prescribing in Australia, researchers and funders should move strategically in future application and evaluation. Recommendations for the consideration of researchers, funders and policy makers include

• A united social prescribing institute that provides opportunities for cross-disciplinary, well-resourced and consistent exploration of social prescribing, both at a policy and practice level.
• High-quality research to explore the impact and reach of social prescribing in the Australian context and encourage researchers to consider the outcome domains identified in this research to enable the amalgamation of data in future comprehensive reviews.
• Longitudinal studies, that enable exploration of outcomes and impact beyond the limit of the social prescribing program [53].
• Support from policy makers and funders to enable consistent exploration of social prescribing across the various funding models in Australia. International literature demonstrates that social prescribing has been used in hospital contexts [54, 55], community-based settings [56, 57], with children [58], people with disability [57] and older adults [38], including veterans [59, 60]. This suggests that opportunities exist in Australia across federally funded health care, state-funded healthcare, the National Disability Insurance Scheme, My Aged Care and Department of Veteran Affairs to explore social prescribing initiatives and evaluation.
• Research investigating the mechanisms of ‘how’ social prescribing works (e.g., the theoretical mechanisms of change), as well as the relational aspects of the approach which may influence how effective it is (e.g., the quality of the relationship with the link worker).
• A core set of outcome domains and measures to support benchmarking is required to be developed by Australian researchers, healthcare workers/referrers, link workers, funding bodies and consumer contribution to ensure alignment with the Australian healthcare landscape.
• The outcome measures across all studies included in this research were selected by the research team. A risk of such an approach is that outcome domains and measures may not hold meaning for participants receiving social prescribing [61]. Consequently, the addition of specific tools that align with the more specific study goals, should also clearly articulate participant consultation in determining an individual’s expectation [62].

Conclusion

In Australia, primary healthcare systems are at capacity, and social prescribing is an internationally recognised approach that has been used to address the health and social needs of individuals. The results of this research indicate that multiple opportunities exist for social prescribing across the broad range of Australian funding models and that both ‘person-level’ and ‘system-level’ domains are critical for social prescribing evaluation. The development of consumer partnerships to develop a core outcome set that can guide both national and international practice and policy is required. Lastly, in addition to the use of specific tools that align with specific study goals, there is a need for future research to clearly articulate participant consultation in determining an individual’s expectation and preferred outcomes of participating in a social prescribing program.

Supporting information

Acknowledgments

The authors acknowledge the contribution of Rachel Ritchie (Librarian, Southern Cross University) for her help in developing the search strategy for this review.