Systematic Review of Contemporary Theories Used for Co-creation, Co-design and Co-production in Public Health

Abstract Background There is a need to systematically identify and summarize the contemporary theories and theoretical frameworks used for co-creation, co-design and co-production in public health research. Methods The reporting of this systematic review follows the Preferred Reporting Items for Systematic Reviews and Meta-Analyses. Given substantial interest in and application of co-creation, co-design and co-production, we searched PubMed, CINAHL, Scopus and APA PsycINFO from 2012 to March–April 2022. A quality assessment and data extraction for theory content was performed. Results Of the 3763 unique references identified through the comprehensive search strategy, 10 articles were included in the review: four articles named co-creation, two articles named co-creation and co-design, two articles named co-production and co-design, and two articles named co-design. Empowerment Theory was employed by two articles, whereas other theories (n = 5) or frameworks (n = 3) were employed by one article each. For the quality assessment, eight articles received a strong rating and two articles received a moderate rating. Conclusion There is little indication of theory applications for the approaches of co-creation, co-design and co-production in public health since 2012, given 10 articles were included in this review. Yet, the theories described in these 10 articles can be useful for developing such co-approaches in future public health research.


Background
3][4] For instance, it can lead to more relevant knowledge production and better tailored solutions.7][8] Hence, this review nominates Vargas et al .'s 7 definitions.
Co-creation is defined as an all-encompassing principle about collaboration and innovative problem-solving among various stakeholders across all initiative phases (e.g., from problem identification to evaluation).Co-design is distinguished as the active collaboration among stakeholders relating to solution design, given a pre-determined problem, whilst co-production is about engaging stakeholders in the implementation of previously set solutions to an already agreed problem, in prioritizing the optimal usage of available resources.In addition, crucial differences exist in terms of stakeholder roles and the nature of engagement in such coapproaches. 7,9,102][13] This may be partly due to inconsistencies and confusion in how these terms are used, coupled with potential knowledge fragmentation between fields, given the interdisciplinary nature of co-creation.
5][16] Theory can offer more transparency, consistency and rigour in the research process, for example, helping to evaluate impact on co-creators and of the co-created initiatives.Whilst the most extensively cited guidance posits that public health intervention development must be underpinned by theory and an evidence base, 3 understanding the different theories about the application of co-creation approaches, 17,18 across all initiative phases, is key.To our knowledge, there is no systematic review consolidating the contemporary theories used for co-creation, co-design and co-production (in public health research) to date.

Aim
We aim to identify and summarize the contemporary theories/ theoretical frameworks used for co-creation, codesign and co-production in public health research.We used the definitions of 'theory' and 'theoretical framework' from Kivunja 19 for guiding this review.
• Theory covers connected concepts, definitions and hypotheses to depict a systemized understanding of a given phenomenon via the connection to variables, with the aim of both delineating and predicting such phenomena.• Theoretical frameworks provide a structure that encapsulates concepts and theories, which results from historically tested and disseminated knowledge.

Methods
The reporting of this systematic review aligns with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA). 20This review progressed on the basis of a registered protocol (PROSPERO ID: CRD42022324074). 21

Search strategy
3][24] The search included controlled and free-text terms for synonyms of 'theory' and 'co-creation' or 'codesign' or 'co-production', with no restriction for methodology or language.The reference lists of the included articles from the full-text screening were reviewed to identify additional relevant articles.

Selection of articles
Original studies (including protocols) had to fulfil the following criteria: contain an explicit reference to named theory/ theoretical frameworks used for co-creation, codesign and co-production in the public health research.This includes theory use for data analysis/ interpretation relating to comprehending feedback on co-creation, co-design and co-production from the participating stakeholders.Using this review's working definition of public health, the publications were to explicitly use the terms 'co-creation', 'co-design' or 'co-production', where theory/ theoretical frameworks were to relate to any or all parts of its approach or its particular (direct) elements.Given theory/ theoretical frameworks about the content of the intervention or product, for example, behaviour change theories (The Capability, Opportunity, Motivation, Behaviour (COM-B) model; Behavioural Change Wheel, etc.) were excluded, in addition to meta-theories.Grounded Theory, Theory of Change and its synonyms, such as programme theory, were also excluded as rendered a methodology, not a theory. 25,26

Electronic database and reference checking
Duplicate articles were excluded by a medical information specialist (H.K.) using Endnote X20.0.1 (Clarivate), following the Amsterdam Efficient Deduplication method 27 and the Bramer method. 28All references were then uploaded to Rayyan systematic software, 29 to perform screening.
The lead author (K.M.) screened the titles and abstracts of all articles against the criteria.Duplicate screening was performed by one of three co-authors (Q.A., G.L., V.A.K.), each screening one-third of all articles.Next, the full-text screening was similarly conducted using all articles remaining-with the lead author (K.M.) having screened all fulltext articles and three co-authors (Q.A., V.A.K., S.C.) performed double screening.Any differences in screening assessment were solved through deliberation among the reviewers.
No additional references were identified from scanning the reference list of the included articles, as screened by lead author (K.M.) and a co-author (G.L.).

Quality assessment
Two authors (K.M. and M.C. or T.A.) rated the methodological quality of the included studies using the quality assessment tool proposed by Bergeron et al . 30K.M., M.C. and T.A. pilottested the methodologic quality assessment to ensure the following items was interpreted consistently: (i) Is the methodology identified and justified?(ii) Was a theoretical lens or perspective used to guide the study, with a reference provided?(iii) Is the theory/ theoretical framework clearly described?(iv) Is the theory/ theoretical framework easily linked to the problem (i.e.co-creation, co-design or co-production)?(v) If an explicit theoretical framework is used, are the concepts adequately defined?(vi) Are the relationships among the concepts clearly operationalized for the context under study?
These questions necessitate a scoring of 'yes', 'somewhat' (i.e.partial information) or 'no'.The 'yes' responses to the questions were added up as a final score.An overall score of three or less in the 'yes' responses was categorized as 'moderate', and an overall score of 4-6 was categorized as 'strong'.The lead author (K.M.) assessed 100% of the included articles, and two co-authors (M.C., T.A.) each assessed 50% of the articles.

Data extraction: theory analysis
Theory analysis took a modified approach from Nadalin Penno et al .'s work 31 -previously adapted from Walker and Avant. 32The theory analysis was 100% performed by the lead author (K.M.), alongside two co-authors (M.C., T.A.) who each extracted from 50% of the articles.K.M., M.C. and T.A. piloted-tested the theory analysis to ensure the following items were interpreted consistently: (i) Publication year (ii) Country of origin (iii) Whether article pertains to co-creation/ co-design/ coproduction (iv) Name of the explicit theory/ theoretical framework, including any citation reference (v) Rationale for the use of the theory/ theoretical framework (vi) Delineation of how the theory/ theoretical framework is used (vii) Methodological approach and (viii) Study setting/ context

Data analysis
The modified theory content fitting the eight sections of the extraction table of the included articles were presented in summary descriptions.Named theories were categorized based on whether the included article explicitly referenced cocreation/ co-design/ co-production.

Results
Figure 1 shows the PRISMA flow diagram presenting the screening process.
The search strategy identified 427 articles for the full-text review phase.Ten articles fulfilled the inclusion criteria, hence included in this review.

Quality assessment
Eight articles were rated strong and two articles received a moderate rating (Appendix 2).Reasons for a moderate score include not justifying the choice of co-creation/ co-design/ co-production approach, the theory not easily being linked to co-creation/ co-design/ co-production, the concepts not being adequately defined about the use of explicit theoretical framework and the relationships among concepts were not clearly operationalized for the study context.

Summary of studies included
Table 1 displays the findings based on the Theory Content Analysis of included articles.Regarding the country of origin, one article was based in the Netherlands, 33 two were based in both the Netherlands and one other country, Turkey 47 and Hungary, 37 two were based in Denmark, 43,45 one originated from the USA, 35 two from UK, 50,53 one from Scotland (Glasgow) and Australia, 55 and one from Canada. 57our articles referred to co-creation. 33,35,37,43Two articles referred to both co-creation and co-design. 45,47Two articles referred to both co-production and co-design. 50,53Two articles referred to co-design. 55,57ll included studies used a predominantly qualitative study design, in terms of process evaluation of youth participatory research, 33 design-based research, 45 mixed-methods case study design incorporating mainly quantitative methods with a qualitative assessment, 47 survey design, 35 qualitative methodology, 37 participatory learning and action research, 50 ethnographic approach 43,53 and co-design processes 55,57 (e.g., workshops).The studies demonstrate various collaborative approaches across the school and community setting, 33 community of pre-school children, 45 NCD (non-communicable disease) prevention, 47 culturally sensitive intervention development, 35 older adults involved in communication intervention, 37 community-based rehabilitation, 43 end-of-life care, 50 community setting in health service development, 55 health research projects in different settings, 53 community setting regarding caregivers and eHealth tool development. 57able 2 presents the summaries of the named theory/ theoretical framework.

Co-creation
Empowerment Theory was used by two articles. 33,35Anselma et al . 33referenced this theory to Israel et al . 34for evaluating the influence of the youth-led participatory action research (process evaluation), as linked to children's empowerment.Specifically, this theory was used for the interview guide, forming provisional codes for data analysis and evaluation.Cueva et al . 35used the same theory, referencing Perkins and Zimmerman, 36 to inform a culturally respectful intervention to empower individuals and endorse wellness.This study's surveys were developed using such theory for intervention evaluation.
One article 37 used two theories, namely Social Learning Theory and Narrative Theory.It attributes Social Learning Theory to the references of Bandura 38,39 as well as Mar and Oatley 40 and attributes Narrative Learning Theory to Mar and Oatley, 40 Moyer-Gusé 41 and Schank and Abelson. 42These theories were justified as compatible with the applied photo stories method, adopted as a health literacy intervention and applied over five phases starting with a literature review and stakeholder analysis.Furthermore, this article argued that narrative communication employs story structures that present a familiar point of interaction, ensuring the intervention is easier to process.The photo stories method was underpinned by social learning theory, whilst narrative theory was used in the research planning and production.
One article 43 utilized 'Symbolic Interactionism' as their theoretical framework, referenced to Blumer, 44 for Table 2 Theory summaries given by the included articles

Study reference Theory summaries
Anselma et al. 33 Empowerment Theory 34 is about people's ability to gain comprehension and control over personal, social, economic and political aspects in order act towards improving their life.
Cueva et al. 35 Empowerment Theory 36 relates to enhancing wellness and bettering problems, offering opportunities for participants to developing their knowledge and skills, as well as engage professionals as collaborators as opposed to authoritative experts.
Koops van't Jagt et al. 37 Social Learning Theory [38][39][40] regards the notion that observation and modelling mainly influence how and why individuals learn.
Narrative Theory [40][41][42] refers to using story-telling as a structure for a familiar way of interacting, which may lead to increased personal involvement and give users role-models and phase-by-phase scenarios.Thus, it is related to acquiring learning through experiences.
Handberg et al. 43 The Theoretical Framework of Symbolic Interactionism 44 posits that human beings act towards things from personal meaning in life, and that this meaning arises from interactions with others, where such meanings are adapted from an interpretative process used by the encountering individual.
Hoeeg et al. 45 'Social Effectiveness of Interventions' Theory 46 relates to an intervention only being socially effective if it generates a 'shared understanding between researchers and professionals and reconfigures the social relationships between researchers and professionals through processes of "exchange"' (2).
Garton et al. 47 Realist Evaluation (Theoretical) Framework 48,49  Normalization Process Theory 51,52 is a structural approach to comprehending the factors that both serve and restrict implementation.It has constructs around coherence, cognitive participation, collective action and reflexive monitoring.
Clarke et al. 53 Interactional Ritual Change Theory 54 relates to the way 'symbolic meanings, beliefs and norms are transmitted and reinforced through relative stable patterns of social interaction or routine' (236).
Farmer et al. 55 Social Innovation Theory 56 is relayed 'as involving collaborations to co-design and implement solutions to social problems, particularly at local level' (2).Relevantly, co-produced solutions are reinforced to have positive societal effects (for instance, via rising aggregate utilitarian value or empowering stakeholders in innovation processes).
Latulippe et al. 57 Amartya Sen's Theoretical Framework of Social Justice 58 provides an interpretative framework for co-design as a democratic process and as a conversion factor.There are three values for the democratic process: intrinsic value, instrumental value (integrated decision-making) and constructive value as a social construct.Conversion factor aids with making sure that the intervention is inclusive.
disentangling the complexities of rehabilitation with multiple stakeholders and regarding how meaning is co-constructed by stakeholders.This framework was used to form an observation guide and the analyses.

Co-creation and co-design
One article 45 used 'Social Effectiveness of Interventions' theory attributed to Rod et al . 46to comprehend the social effectiveness of interventions.This theory was employed as an analytical method in co-creation processes and exploring design-based research methodology.
In addition, a specified theoretical 'realist evaluation' framework was referenced to Tilley and Pawson 48,49 by one article 47 and was used to account for the strategic changes over time in regard to NCD prevention.It was used to demonstrate the sense-making process that occurs between context, mechanism and outcome patterns of a programme.

Co-production and co-design
Duke et al . 50referred to Normalization Process Theory, referenced to May, 51,52 which was substantiated by combining participatory learning and action research-believed to retain a positive influence over the design quality of the intervention and implementation, as enabling diverse knowledge and expertise.This theory was used in the pilot implementation phase, as relating to the data collection.
Clarke et al . 53referred Interactional Ritual Change Theory to Goffman and Collins. 54They adopted this theory as pinned to the importance of enabling inclusivity of stakeholdersby highlighting everyday routines/ rituals in relation to more sense of belonging.In addition, the authors suggest that this theory offers a novel perspective regarding situated coproduction practices, which was used for the data analysis in exploring how co-production is applied to promote sustainable and inclusive initiatives.

Co-design
Social Innovation Theory, used by Farmer et al . 55is referenced to Ayob et al . 56to consider initiatives co-designed with communities as social innovation, identify helpful initiatives in co-design approaches and aid change in public health.It was also used for exploring how top-down community participation processes permit diverse participants to exchange, for instance, knowledge in new contexts.
Latulippe et al . 57used Amartya Sen's theoretical framework of social justice 58 as informative for co-design in exploring how it enables an inclusive intervention and can lead to the formation of a universally accepted tool (e.g., eHealth tool), for benefitting future users (e.g., caregivers of functionally dependent older people).

Main finding of this study
Since 2012, 10 articles explicitly used theories for co-creation, co-design and co-production in public health research, comprising six distinct theories and three distinct theoretical frameworks.Empowerment theory was utilized by two different articles, whereas the other theories were used once per article.The 'how' and 'why' of all theories were outlined, including theory summaries.For instance, most theories used by these articles had informed the methods (focus groups, etc.), with their relevance justified.

What is already known on this topic?
3][4] Yet, it is evident that the conceptualization behind these co-approaches requires more development. 59Specifically, there is a lack of clarity and delineation about how these terms are understood and applied in public health research (and other fields).In turn, this risks knowledge fragmentation and concept stretching.[13] What this study adds Our findings can enable theoretically informed public health research applying co-creation, co-design and co-production.For instance, Empowerment Theory can inform how stakeholders can be empowered, as linked to outcomes 33 and explicit aims. 35Some of the same named theories from the public health literature fit with that of other fields that use co-creation, co-design and co-production.For instance, Empowerment Theory was used as a method to understand the concept of customer empowerment and thereby inform about knowledge value co-creation 60 ; the Theoretical Framework of Symbolic Interactionism was used to analyse the social interactions between stakeholders involved in a forest management conflict, for instance, as this framework is useful for identifying ways in which role confusion contributes to conflict 61 ; and the Normalization Process Theory was used in health services research to support the implementation of complex interventions, by recognizing sense-making, engagement, action and monitoring-and adding to its transparent evaluation. 62][65][66] This theory may be of closer relevance to co-creation given empowerment is about the importance of agency to influence change.Other theories were only used once which may be because co-creation research in the field of public health is relatively new, albeit evolving.We encourage research to be directed towards all the named theories covered by this review, including whether Empowerment Theory is more relevant than other theories for co-creation.In addition, we recommend future researchers to give clear definitions for creation, co-design and co-production to permit conceptual clarity, and thereby give more strength to the interpretation of its relevant use of theory.

Limitations of this study
Implicit theories utilized for public health research for cocreation, co-design and co-production may have been missed as this review was solely focused on named theories.The same applies for studies using alternative terms for co-creation, codesign and co-production that incorporated explicit theories in public health research.Furthermore, the articles naming an approach (e.g., co-creation) may not have utilized the same definition of the term as this review.However, our search strategy was optimized through reference list checking and by trialling different search terms for results comparison.The quality assessment tool did not cover 'low' score ratings, potentially giving a more positive impression about studies than warranted-yet this rating did not influence our theory analysis.

Conclusion
Currently, there is a lack of explicitly named theory supporting co-creation, co-design and co-production in public health research.Given a wealth of interest and application geared towards co-creation, co-design and co-production research, more progress needs to be made in relation to the incorporation of named theory/ theoretical frameworks informing these co-approaches.This review helped to identify some usable theories and theoretical frameworks, in addition to delineating how and why the theories were used for such coapproaches.Thus, public health researchers can contemplate the differing applications and justifications for the use of such theories and theoretical frameworks in this way.

Originality
This article is an original piece of work which has not been published before.

Table 1
Theory Content Analysis of included articles, sorted by referenced explicit term (co-creation/ co-design/ co-production), name of the theory/ theoretical framework and study reference Fig. 1 PRISMA flow diagram.

Appendix 2
Quality assessment extraction table