To understand the heterogeneity of research traditions that study urban health, we apply the meta-narrative approach to systematic literature review. A meta-narrative review is a type of systematic literature review that is designed for topic areas that are researched by diverse research traditions, with different conceptualisations and methodologies (10–13). Through an explicit, rigorous and transparent process, meta-narrative reviews identify, articulate, synthesise and interpret a diverse body of literature in a topic area (11, 14, 15). In this review, we will systematically collect and analyse the literature that addresses the issues concerning the impact of urban built environments on human health and seek to make sense of the complex and contested knowledge in this topic area.
The protocol for this review was developed in accordance with the RAMESES (Realist And MEta-narrative Evidence Syntheses: Evolving Standards) publication standards (15) and quality criteria suggested in the associated meta-narrative review training materials (14). Because meta-narrative reviews are different from traditional systematic literature reviews that they are designed to reflect the heterogeneity of the research methodologies, the protocol is not fully compatible with the Preferred Reporting Items for Systematic Review and Meta-Analysis Protocol (PRISMA-P) guidelines. Nevertheless, we have populated the PRISMA-P checklist as it still provides critical value to the systematic review process [Additional file]. The protocol is registered with PROSPERO [CRD42020192992].
The six guiding principles of the meta-narrative review (15) are integrated into the review process as articulated below (Table 1).
Table 1
Meta-narrative review principles and applications
Principle
|
Definition
|
Application in this review
|
Pragmatism
|
The review should address what will be most useful to the intended audience
|
The objective of this review is to understand the main paradigms in urban health. In a transdisciplinary field of research and practice, articulating the non-overlapping characteristics of different paradigms is critical to attain coherence and collaboration across disciplines, sectors and paradigms.
|
Pluralism
|
The topic should be illuminated from multiple angles and perspectives
|
We explore the current knowledge base in various disciplines, including public health, urban planning, local/city governance and urban studies. A list of relevant disciplines and journals will be drafted to utilize for the hand selecting of literature to avoid any exclusion of the disciplines.
|
Historicity
|
The deepest understanding of a topic comes from studying its evolution over time
|
The genealogy and clusters of the literature will be analysed using bibliometric methods. Landmark documents will be recorded and traced to study the evolution of the paradigms.
|
Contestation
|
Conflicting data from different research traditions should be examined to generate higher-order insights
|
Differences between the conceptualisation of urban health, causal pathways, methodological approaches and policy solutions will be highlighted. Details on the application of this principle will be explored further in the data extraction, analysis and synthesis phases.
|
Reflexivity
|
Reviewers should continually reflect on the emerging findings
|
The protocol will be updated to reflect the changes to the process as findings emerge. Any changes made to the review that were initially planned will be described and justified in the final report.
|
Peer review
|
Emerging findings should be presented and discussed with an external audience
|
The emerging findings will be communicated with peers via individual consultations with experts and presentations at conferences and meetings. A website will be developed as a platform for the wider community to engage in the process as well as for dissemination of the findings.
|
The methods section of this protocol is presented according to the six phases of a meta-narrative review recommended by the RAMESES publication guideline.
Planning phase
A review team consisting of the authors of this protocol was assembled, who developed the research questions and drafted the protocol. The review team provided feedback on the draft protocol. Knowledge users, defined as the broader network of researchers and practitioners in the field of urban health, will be approached to serve as external expert panel members for the review and will be consulted if additional assistance is needed.
Search phase
Eligibility criteria
Publications to be included in the review will be limited by language (English) and publication types (journal articles, reviews, books, book chapters, editorial and opinion pieces and reports). All study designs, including empirical and non-empirical studies, and all publication years will be considered for inclusion. The topic of the paper must explicitly focus on urban human-constructed physical environments and human health at the city or urban scale and must address one or more of the conceptual, theoretical, methodological or instrumental dimensions on this topic.
Search strategy
The main objective of the search is to collect a comprehensive list of the literature on the topic area to capture the diversity of the urban health research traditions and paradigms. The publications will serve as primary data source to analyse the meta-narratives of each paradigm. The balance between comprehensiveness and precision of the search is resolved by applying the concept of saturation as the criterion, a concept borrowed from qualitative research methods. Saturation is achieved when no new information is generated by collecting additional data. Because this review is a ‘knowledge-building and theory-generating’ type of systematic review, there is no intrinsic value in continuing the search unless there is additional theoretical contribution (16). This is in contrast to those reviews which study the aggregation or summation of concepts, where ‘the more the better’ approach is preferred. The attainment of saturation will be determined at the appraisal or synthesis phase when the review team observes a conceptual saturation of the findings from the identified literature, and decide that the addition of unidentified studies will only contribute to marginal changes to the findings. (17–19).
The search will take three main strategies – (a) a double-sided snowballing search, (b) a search in electronic databases using search terms, and (c) an additional hand search. The double-sided snowballing search will include a forward search of all papers that cite the landmark work identified in the mapping phase, and a backward search that collects the literature included in the reference list of these papers (20). A search using keyword search terms will be conducted in relevant multidisciplinary scientific databases (e.g. ProQuest, Web of Science (Science and Social Sciences Citation Indices), Scopus) and topic-specific databases (e.g. Urban Studies Abstracts – EBSCO, Medline – Ovid, Embase _ Ovid). The search terms will include those related to city (“city” OR “cities” OR “urban” OR “local” OR “municipal”) combined with concepts on the built environment (“built environment” OR “physical environment” OR “infrastructure” OR “planning” OR “design”) and search terms related to health and health equity. An additional hand search of key journals and publications by key organisations will be conducted to maximise comprehensiveness.
Selection process
The publications identified from these three strategies will be compiled in EndNote and exported to Covidence to be screened for inclusion in the review. Two reviewers will screen the title and abstract of each publication to decide inclusion in the review. Any disagreement will be resolved by consensus. Details of the inclusion criteria will be added and refined as the reviewers proceed with the screening. It is also expected that the search strategy will be iteratively revised by the paradigms identified in the mapping phase.
Mapping phase
Mapping urban health paradigms and defining parameters
As mentioned earlier, we start with four paradigms on urban health – the medical-industrial city, urban health science, healthy built environments and health social movement - as the initial paradigms for this review (Table 2). These four paradigms explicitly address the relationship between the urban built environment and human health, and each has a set of conceptual, theoretical, methodological and instrumental dimensions. Findings from the search and discussions with the wider expert community may introduce additional paradigms to the review. For example, there may be additional paradigms in the fields of environmental health, spatial justice, or civil engineering which were not covered in our initial mapping. Alternatively, the initial paradigms may merge or subdivide as more information is added from the review process.
Table 2
Four initial urban health paradigms
|
Medical-industrial city
|
Urban health science
|
Healthy built environments
|
Health social movements
|
Main concept
|
The healthcare industry is a main driver for urban growth and development.
|
Production of evidence that reveal and confirm causal relationships between the built environment and health.
|
Influencing the planning system to integrate health to urban planning decisions.
|
Empowering the community and value-based decision-making for healthy cities.
|
Related disciplines and fields
|
Business, urban planning, development, economics
|
Public health, epidemiology, population health, implementation science
|
Urban planning, urban design, land-use planning
|
Health promotion, community development, social justice, sociology
|
In the mapping phase, we will develop a set of parameters for each paradigm. For example, we will define the characteristics of each paradigm based on their definitions of urban health, the theoretical frameworks that explain the relationship between urban physical environments and health, the methodologies to research (e.g. what is counted as knowledge, evidence, etc.) and the solutions such as the policies, practice and solutions to improve urban health. In particular, we will include the concepts of health equity and the consideration of power in decision making in reviewing the parameters. These parameters will be used as a guide to search for data on the dimensions of the multiple paradigms in each publication. We will apply the set of parameters to assign each piece of an included publication to its corresponding paradigm(s). To ensure high inter-coder reliability, the review team will first code a small sample of the publications and any conflicts or disagreement will be examined and discussed. Any disagreement will be resolved by consensus within the review team.
Bibliometric analysis to visualise network and clusters
A supplementary bibliometric analysis using the dataset of the final set of articles will be conducted to map the genealogy of citations and the author network analysis (21). The findings will visualise clusters of researchers and relationships between publications. These visualisation data will provide information to triangulate the different paradigms and research traditions.
Identifying landmark works
The mapping phase includes identifying landmark works that formed the foundation for the paradigms and are recognised by scholars in the field as highly influential in shaping subsequent research and practice (11). They can be conceptual papers or reports, or empirical studies that formed a model for future work in the paradigm. We will triangulate this with the citation metrics data and the findings from the bibliometric network analysis.
The following inclusion criteria will be applied to identify the landmark sources (11)
-
Is the paper part of a recognised research paradigm, that is, does it draw critically and comprehensively upon an existing body of scientific knowledge and attempt to further that body of knowledge?
-
Does the paper make an original and scholarly contribution to research into the topic area?
-
Has the paper subsequently been cited as a landmark contribution (conceptual, theoretical, methodological or instrumental) by competent research in that tradition?
-
Is the paper an exemplar of a recognised research paradigm and its parameters?
The review team will independently score and nominate landmark sources according to the above criteria. Discussions will be held with external experts to attain consensus.
Appraisal phase
In the appraisal phase, we will extract data from selected publications by coding the conceptual, theoretical, methodological and instrumental concepts. The main outputs from this phase include a codebook with the descriptions of the codes, an NVivo project with coded data of the included literature, development of the quality assessment criteria for each paradigm, and the quality assessment of the literature.
Data extraction
First, a code system (or a data extraction template) will be developed based on the parameters of the paradigms developed in the mapping phase. The code system and data extraction template will include:
-
Bibliometric meta data (e.g. author, publication year, title, type of publication, etc.)
-
Research questions and how they were framed, and conceptual and theoretical issues;
-
Preferred methodologies, study designs and quality criteria;
-
Key actors (e.g. leading scientists or commentators) and events (e.g. conferences) in the unfolding of the tradition;
-
Landmark empirical or theoretical studies;
-
Significant findings and how they shaped subsequent work;
-
Key debates and areas of dispute within the tradition, including links with or breaches from other traditions.
-
Characteristics in cross-disciplinary approaches (interdisciplinary, multidisciplinary or transdisciplinary).
Details of the coding system will be determined by the review team. Whether data extraction will be based on the abstract or the full-text will be determined at a later time, after the review team learns more about the body of literature. We will use the NVivo qualitative data analysis software to efficiently organise the data. Using NVivo software for qualitative coding will also allow us to refer back to the original data and transparently track the collaborative process. The reviewers will independently extract data and the coded data will be examined to ensure inter-coder reliability. All data will be stored in the approved research data storage system provided by the lead author’s institution and handled in accordance with the institution’s data management standards and guidelines.
Quality and risk assessment
Each paradigm will inherently endorse a different set of standards for assessing the quality and risk of bias of studies. Criteria to assess the quality and risk of bias will be taken from the paradigms included in the review, particularly from the landmark papers that have been accepted by the paradigm as authoritative. The publications, now classified to one or more paradigm, will be assessed against the corresponding quality criteria.
Synthesis phase
Building the meta-narratives of each paradigm
Synthesis involves comparing and contrasting the meta-narratives among the different paradigms to identify and compare how they have conceptualised the topic, how they have theorised it, and the methodological approaches and study designs used. After coding the data from the primary studies within each paradigm, we will construct a meta-narrative for each paradigm, and then compare and contrast the paradigms.
Comparing meta-narratives across paradigms
Synthesis across paradigms may occur at a high level of abstraction and may involve one or more of the following :
-
Paradigm bridging (seeking commonalities in conceptual and theoretical assumptions)
-
Paradigm bracketing (highlighting differences in these assumptions)
-
Interplay (exploring tensions)
-
Meta-theorizing (exploring patterns that span conflicting understandings)
The key questions we ask in this phase include : (1) What is the range of questions the paradigms address across the four dimensions? (2) What are the commonalities and conflicts of research findings across the paradigms and how can the discrepancies be explained? (3) What are the overall key findings and implications? (4) What are the main gaps and where should future research be directed?
Recommendation phase
The final phase of the review includes drafting the final report with key messages and recommendations for practice, policy and further research. The final report will be developed through reflection and discussions with the review team and feedback from the wider urban health epistemological community. Reporting will be in compliance with the RAMESES publication standards (15). Any changes made to the protocol will be documented in the final review report.
A dissemination strategy will be developed to further communicate with the broader knowledge user community. Components of the review (e.g. the protocol, methodology, meta-narrative review findings) will be reported through traditional academic communication channels such as academic journal publications and conference presentations. Presenting the findings of the review over the various stages will provide a form of triangulation to ensure the validity of the review and address some of the issues occurring from meta-biases such as publication bias or selective reporting. The team will further identify a broad range of potential knowledge users and stakeholders and develop innovative strategies to effectively communicate with them.