Co-Produce, Co-Design, Co-Create, or Co-Construct—Who Does It and How Is It Done in Chronic Disease Prevention? A Scoping Review

Co-production in health literature has increased in recent years. Despite mounting interest, numerous terms are used to describe co-production. There is confusion regarding its use in health promotion and little evidence and guidance for using co-produced chronic disease prevention interventions in the general population. We conducted a scoping review to examine the research literature using co-production to develop and evaluate chronic disease prevention programs. We searched four electronic databases for articles using co-production for health behaviour change in smoking, physical activity, diet, and/or weight management. In 71 articles that reported using co-production, co-design, co-create, co-develop, and co-construct, these terms were used interchangeably to refer to a participatory process involving researchers, stakeholders, and end users of interventions. Overall, studies used co-production as a formative research process, including focus groups and interviews. Co-produced health promotion interventions were generally not well described or robustly evaluated, and the literature did not show whether co-produced interventions achieved better outcomes than those that were not. Uniform agreement on the meanings of these words would avoid confusion about their use, facilitating the development of a co-production framework for health promotion interventions. Doing so would allow practitioners and researchers to develop a shared understanding of the co-production process and how best to evaluate co-produced interventions.


Introduction
In Australia, as in many other high-income countries, chronic diseases place a significant and persistent burden on the community, with the social and economic consequences having a detrimental effect on an individual's quality of life [1]. Much of the burden caused by chronic disease is preventable, with modifiable risk factors including tobacco use, overweight and obesity, physical inactivity, and unhealthy dietary behaviours [2]. Governments are often the lead agency for the development and the implementation of health promotion strategies and programs to prevent health behaviour-related chronic disease at the population level, as well as research into their effectiveness as a way of promoting healthy choices, preventing disease, and keeping people out of hospital [3,4]. 'Co-production' is a mechanism whereby 'stakeholders' (identified as including end users or intervention target audience, health researchers, academics, policy and practice partners, decision makers, and funding representatives) can collaborate, generate relevant knowledge, and apply it to practice [5][6][7]. It involves key stakeholders in the development of

Materials and Methods
We conducted a scoping review to identify the nature and extent of co-produced interventions addressing chronic disease prevention [19]. A scoping review was chosen because its purpose is to identify the types of evidence in a research field, clarify key concepts, explore how research is conducted on a topic, and identify gaps in the literature [20]. The methodology was guided by an established framework for scoping studies [21,22]. The Preferred Reporting Items for Systematic reviews and Meta-analyses extension for Scoping Reviews (PRISMA-ScR) checklist was used to ensure that this scoping review achieved quality standards of practice and reproducibility [23].
The development of the search strategy was iterative. First, we conducted a preliminary search of the literature, which included the term 'co-production' to identify other terms related to co-production that were commonly used in the literature [6,14]. These terms were: co-design, co-create, co-construct, partnership, and collaboration. A further literature search was conducted to identify a sample of articles using 'partnership' or 'collaboration' in the development and evaluation of chronic disease prevention programs. Two authors (B.M. and B.O'H.) screened the full text of these papers to determine how partnership (n = 16) and collaboration (n = 12) were used in these studies. We found articles using the terms 'collaboration' or 'partnership' did so to loosely refer to intersectoral (including academic, government, and financial) relationships. The studies screened did not define or describe what was meant by partnership or collaboration in sufficient detail in relation to the development or evaluation of programs with a chronic disease prevention focus for these terms to be included in this study. This observation is supported by Johnston and Finegood [24], who acknowledge the ambiguous and interchangeable use of the term partnership with other terms including collaboration. Therefore, 'partnership' and 'collaboration' were excluded from the final search strategy as they were concepts considered broader than co-production.
We systematically searched electronic databases for peer-reviewed literature (Medline via Ovid, PsycINFO via Ovid, Cinahl, Scopus, and PubMed) for articles using a co-production, co-design, co-creation, and co-construction approach to achieve a lifestylerelated health behaviour change [6,14]. The search was conducted in May 2020 and updated in March 2021. Search terms included a combination of medical subject headings (MeSH) terms and keywords as outlined in Table S1. Reference lists of all included studies and relevant interventions known to the authors were searched for additional studies. We included English, peer-reviewed articles with studies reporting protocols and/or outcomes of interventions (primary and secondary prevention) using a co-production, co-design, co-creation, and co-construction approach in any health promotion setting. The outcomes of interest were chronic disease prevention-related behaviours outcomes such as smoking, physical activity, diet, and/or weight management. The populations of interest included individuals, communities, or populations at risk of developing health behaviour-related chronic disease. We did not limit studies by research design or publication year. Studies of interventions with a clinical orientation (e.g., those targeting service delivery, rehabilitation, or medication adherence) were excluded.
Search results were combined in EndNote X9.3.3, and duplicate references were removed. Three authors (B.M., B.O'H., and L.C.) independently scanned titles and abstracts to determine inclusion eligibility. The full text of eligible papers was independently reviewed by B.M., B.O'H., and L.C. according to the pre-determined inclusion criteria outlined above, with discrepancies resolved by discussion. Study characteristics extracted included how study authors defined the 'co-words' used, who was involved in the co-production, the process and impact measures to evaluate interventions using co-production, and the sections of the paper where co-words were mentioned. Data were further tabulated according to the purpose of the study, the prevention focus, the target population, and the collaborative technique used in relation to co-production (Table S2).

Study Selection
Database searches identified 589 publications, and an additional two publications were identified by citation searching. Following duplicate removal and title and abstract screening, 117 full-text articles were reviewed for inclusion eligibility, resulting in the inclusion of 71 articles (64 unique studies and 7 reviews) ( Figure 1).

The Operationalisation of Co-Production in the Development and Evaluation of Chronic Disease Prevention Interventions
Co-design was reported in 39 articles (including 4 reviews) [ [30,64,67,86]. All articles that defined co-design, plus an additional article with a cocreation focus [68], identified that co-design aligned with a participatory design approach, in which end users or stakeholders of intervention are engaged in the research process. As one of the papers stated, the process of co-design is illustrated as a 'golden thread' running through all stages of public health research [84], enabling the contributions of end users to be incorporated from intervention development and testing to implementation and dissemination [63,82,87]. In particular, the aim is to empower stakeholders as part of the design process by recognising their expertise in their own experiences [53,69]. Intervention co-design was described in these studies as an iterative and creative collaboration or partnership between end users and relevant stakeholders and intervention designers [53,63,84,87]. While not all articles explicitly defined what they meant by co-design, collaboration between multiple stakeholders as part of the co-design process was described in the methods in the majority of papers reporting co-design [25,53- [27,36,[39][40][41]47,[49][50][51]56,73,75], and despite using the term within their paper, five did not clearly outline what was meant by cocreation in the study [42,43,46,62,94]. In those papers that did define it, co-creation was defined as an active process between people with shared goals but different expertise and skills, by which stakeholders were enabled to be directly involved in the generation of an intervention or solution [38,45,48,68,72,85]. There were similarities in how co-creation was defined to the definitions of co-design described above. Namely, co-creation is described as having developed from participatory design [85] and the collaborative engagement of all stakeholders, including end users, throughout the process of developing and implementing an intervention [38,48,68,72,85]. Two articles described the inclusion of co-creation in the development of behaviour change interventions as (a) reducing barriers to change [56] and (b) providing insights into motivation for change [51].
Co-production was reported in 16 articles (13 unique studies and 3 reviews) [13,25,28,29,[31][32][33][34][35]37,52,62,[88][89][90]94]. Two studies provided an explicit definition of co-production [33,37], eight described the process by which co-production was used to develop or evaluated their intervention [13,28,29,32,34,35,52,90], and six did not provide details on what they meant by co-production [25,31,62,88,89,94]. Similar to co-design and co-creation, coproduction was defined as involving the target audience in the design and implementation of an intervention [37]. The process of co-production, according to these articles, incorporated the implementation, stakeholder, and participant contexts into the intervention development, implementation, and evaluation [32]. As with co-design and co-creation, co-production was described as using participatory approaches to involve participants and stakeholders in an equal and reciprocal relationship [35,52] for the iterative development of an intervention [13,28,29]. The authors felt that such a process gave rise to services that meaningfully met the needs of individuals and communities, and represented an engaging process to achieve behaviour change among end users [33].
Co-construction was used in three unique studies [44,57,62]. None defined co-construc tion but described the process as involving collaboration with stakeholders and end users through all stages of intervention development and evaluation [44,57,62].

Evaluation of Chronic Disease Prevention Interventions Developed Using Co-Production
Ten studies included mention of the acceptability [26,31,51,60,92] and feasibility [26,27,31,36,47,51,58,60,73] of implementing a co-produced intervention. Among these studies, there was no consistency in the way acceptability and feasibility were measured. Most authors concluded that a co-produced intervention was feasible because the views of the target audience were able to be incorporated into a revised intervention. Similarly, acceptability was determined by implication because of the acceptability of the target audience's involvement or as measured post-intervention development through questionnaires or qualitative interviews. Three protocol papers included plans to undertake a process evaluation, with some consideration given to issues of implementation of a co-produced intervention [32,35,77].
Eighteen studies reported on the evaluation of their intervention in terms of process evaluation (n = 5) [43][44][45]50,92], impact evaluation (n = 9) [47,57,[65][66][67]69,71,78,79] and both process and impact evaluation (n = 4) [29,52,56,76]. These studies did not include an analysis of whether the use of co-production afforded any implementation or outcome advantage and, as such, did not report on the impact that co-production had on the associated implementation or outcomes of the program. A further two-thirds of the papers (65.3%, n = 47) included in the review limited their scope to describing only the techniques used for undertaking co-production, with no evaluation results reported of any kind.

Discussion
This scoping review found 71 articles that reported using co-production when developing a chronic disease prevention intervention or program, with the majority published in the last three years. Our findings highlight that different 'co'-words were used interchangeably within and across many studies, and little attention was paid to whether there were any differences (subtle or otherwise) in their intended use and meaning. The 'co'-words used included co-produce, co-design, co-create, co-develop, and co-construct, either singly or in combination. Although we initially focused on co-production, co-design and co-creation were more commonly used in practice in the selected primary studies. Occasionally, a 'co'-word was used only in the abstract, perhaps as a way of drawing attention to the article, but the body text included no further exploration of the term.
Across the different terms, in the studies we reviewed, 'co-' words were used to describe a process of engaging with the target audience of end users or intermediaries (e.g., health promotion and health practitioners, etc.) [95] of an intervention in a participatory fashion [17]. There were no substantive differences in meaning between co-design, cocreate, co-produce, and co-construct and how they were deployed in reporting on an intervention. There were also no notable differences in the methods used in co-production based on the 'co-' word used by the study. Overall, co-production constituted a formative research process [96], including focus groups, interviews, and other methods of information collection. Again, the literature would benefit from clarity as to whether the different terms are or should be linked in some way to particular techniques.
Through our analysis, it became apparent that those who use a co-production method choose the relevant stakeholders to be involved in the design, implementation, or evaluation of a chronic disease prevention intervention. The most common co-production participants were academics and the target audience, followed by intervention designers, implementers, and policymakers. This finding is not unexpected given that the majority of studies in our review reported on co-production in terms of the development of an intervention rather than the implementation and evaluation. It is uncertain whether there would be a benefit in attempting to define the group(s) to be involved in co-production processes, as this may vary widely with the project and context. More important is to examine the impact of including different groups on the outcomes and implementation of an intervention.
As noted above and by other authors [18,96], our review confirmed the paucity of evidence that examines the impact or effectiveness of co-production processes in chronic disease prevention interventions. This is unsurprising given that most studies included in this review outline the co-production technique, and the few that reported on evaluations used a pre-post design in relation to the interventions' target behaviours. A few studies noted that using a co-production method was acceptable and feasible because a) the studies had been effective at incorporating the views of the target audience in the design of the intervention and b) post design, the intervention had been used by the target audience almost as a proxy measure of acceptability. Future studies should formally evaluate the perceived acceptability and feasibility of co-produced interventions within target populations rather than relying on proxy measures. The review papers we included also reported that studies were more likely to report on feasibility and acceptability rather than impact of the co-production process on intervention outcomes, with Eyles et al. [87] concluding that "sufficiency of reporting was poor, and no study undertook a robust assessment of efficacy" (p 160). Future studies with robust evaluation designs are needed to evaluate the effectiveness of co-produced health promotion interventions so the impact of the co-production method can be determined.
Our findings suggest that there would be merit in developing conceptual or definitional guidance as to what these words mean or include in the chronic disease prevention setting and whether there are differences in meaning or whether they can be used interchangeably to describe the same process. Our review supports the notion that 'co' is suggestive of a co-operative, collaborative, or participatory design [97], as noted by Blomkamp [98], but it is not possible to suggest from our findings any definite differences in meaning between the various 'co'-words. There may be merit in developing a framework that provides greater understanding of the distinctions between various terminology. This could be progressed by borrowing from the health services [99,100] and public administration co-production literature [14,15] which provides some guidance in defining the most used co-words by articulating their differences and then providing a hierarchy of meanings that can be used to guide co-production in chronic disease prevention. Our research also suggests that starting points could be: defining terms by those involved in the collaborative process [6] or using a staged approach to co-production as mapped against a program design cycle [6,13]. There is also an opportunity to explore how co-production does or does not align with community-based participatory research [101] or participatory action research [102], particularly in relation to where along the intervention design and evaluation continuum it fits and also which stakeholders it involves.
This review provides an initial step in overviewing a growing field of research that is 'messy'. While research co-design in health has been included in previous reviews [103], our review is novel in its focus on the use of co-production in the development and evaluation of co-produced chronic disease prevention interventions that aim to change a lifestyle behaviour. A strength is the inclusion of all 'co-' words used in publications and the broad view taken to recognise similarities or differences in their use.
A number of limitations that may affect how the findings are interpreted need to be acknowledged. Only studies published in English were included, potentially excluding relevant studies published in other languages. The search was limited to peer-reviewed literature and did not include a grey literature search. It is possible that policy statements and reports relevant to co-produced chronic disease prevention interventions could have contributed to the review findings. Future reviews should include studies in other languages as well as from the grey literature. The wide range of study designs and research methods are drawn from for this review limited the options available for drawing conclusions for defining 'co'-words and suggesting frameworks appropriate to health promotion interventions [21]. We did not conduct a quality appraisal of the primary studies included, which is consistent with scoping review methodology [19,21] but leads to a broad range of included material. The advantage of considering the breadth of literature is that it provides a structured overview of the current use of co-production in health promotion and provides direction for the focus of future research.

Implications for Practice and Research
Our review suggests that, as with co-production more broadly, co-produced health promotion interventions that aim to prevent chronic disease are not well described or robustly evaluated. The public health literature does not currently provide insight into whether co-produced interventions achieve better outcomes than those that are not coproduced [104]. Co-production, co-design, co-creation, and co-develop seem to be used interchangeably to refer to a participatory or collaborative process involving researchers, stakeholders, and end users involved in the development or evaluation of such interventions. Uniform agreement on the meanings of these words would avoid confusion surrounding their use and facilitate the development of guidelines and/or a co-production framework specific to health promotion interventions. Doing so would allow researchers to develop a shared understanding of the co-production process and how best to evaluate co-produced interventions.