PROSPERO International prospective register of systematic reviews
A systematic review of interventions to boost social relations through improvements in community infrastructure (places and spaces)
Anne-Marie Bagnall, Jane South, Salvatore Di Martino, Kris Southby, Gerlinde Pilkington, Andrew Pennington, Rhiannon Corcoran
Anne-Marie Bagnall, Jane South, Salvatore Di Martino, Kris Southby, Gerlinde Pilkington, Andrew Pennington, Rhiannon Corcoran. A systematic review of interventions to boost social relations through improvements in community infrastructure (places and spaces).
Available from http://www.crd.york.ac.uk/PROSPERO/display_record.asp?ID=CRD42017071098
Review question 1: How effective are interventions designed to improve community infrastructure (places and spaces) in improving social relations and/ or community wellbeing?
- What interventions to improve community infrastructure have been evaluated with regard to social relations and/ or community wellbeing?
- In which settings have interventions to improve community infrastructure (places and spaces) been evaluated with regard to social relations and/ or community wellbeing?
o Is there an association between setting and:
? type of intervention,
? outcomes measured and
? direction and size of effect?
- Are there differences in effectiveness across population groups, particularly those at risk of health inequalities? (for example, people from different socio-economic backgrounds, ethnicity, age or gender)?
- Are there differences in effectiveness across different types of interventions?
o are there differences across interventions and initiatives that have been explicitly planned by agencies (e.g. play areas), and those that have developed informally (e.g. café as meeting place), sometimes called “third spaces”?
o what is the evidence about the effectiveness of interventions within estate regeneration schemes, other neighbourhood or high street renewal schemes, and new housing developments?
- Are there differences in effectiveness across interventions that:
o (i) aim to mix population groups (e.g. intergenerational connections; different ethnicities; community cohesion);
o (ii) are open to a mix of population groups, although this is not an explicit aim;
o (iii) are targeted towards specific population groups, such as those at risk of social exclusion and/ or health and wellbeing inequalities, or are intended to strengthen bonds within a population?
Review question 2: What factors (positive and negative) affect the implementation or effectiveness of the interventions?
Review question 3: What are people’s subjective experiences of interventions designed to improve infrastructure (in relation to social relations and community wellbeing)?
o Do these differ across settings, intervention types, population groups?
o How involved are local communities in design, delivery and evaluation of interventions, and does this influence effectiveness?
As a result of initial scoping searches, we will search the following databases:
PsycInfo, MEDLINE, CINAHL, Social Policy and Practice (covers Social Care Online and Idox), Social Sciences Citation Index, Academic Search Complete, Leisure Tourism, Hospitality and Tourism Complete, Avery Index, GreenFiles and Urban Studies Abstracts. We may also search systematic review databases (i.e. Cochrane database of systematic reviews, DARE, Campbell Library, DoPHER (EPPI-Centre), Joanne Briggs Institute, Epistemonikos), ASSIA and Sociological Abstracts.
We will also search for ‘grey’ literature through Opensigle, topic experts (i.e. review advisors, and contacts through the What Works Centre for Wellbeing) and relevant websites.
A call for evidence will be issued by the WWC-WB, shared on social media and distributed to a mailing list of over 1200 academics and practitioners who expressed an interest in evidence on community wellbeing during the Voice of the User stakeholder engagement phase of the Community Wellbeing Evidence Programme.
Reference lists of key systematic reviews and included studies will be scanned, and citation searching will be carried out for included articles where possible.
An audit table of the search processes will be kept, with date of searches, search terms/strategy, database searched, number of hits, keywords and other comments included, in order that searches are transparent, systematic and replicable as per PRISMA guidelines. The results of the searches will be downloaded into Endnote reference management software.
Types of study to be included
We will include quantitative and mixed methods studies which use experimental designs, and also process evaluations and qualitative studies that relate to the intervention specified above.
We will exclude articles which provide only descriptive information or commentary.
Condition or domain being studied
The aim of this systematic review is to synthesise the available evidence, and describe the quality of that evidence, in relation to interventions that improve or create the community infrastructure that impacts on social relations and/ or community wellbeing. For this review, we are defining community infrastructure as the physical places and spaces where people can come together, formally or informally, to interact and participate in the social life of the community. We intend to produce a synthesis which is accessible and will inform practice and future research in the area.
We will include literature relating to community infrastructure: places and spaces for any community. We will focus on evidence for adults (loosely defined as aged between 16 and 65, but will accept other definitions as presented in studies). If included studies also present evidence relating to other age groups, we will include this where possible, particularly if there is any data on intergenerational relations.
We will exclude studies that include only older adults (as defined by the study authors) or only children (as defined by study authors), as these fall within the remit of two other What Works Centres (the Centre for Ageing Better and the Early Intervention Foundation). We will include interventions aimed at families, such as children’s play areas.
We will include studies which have been carried out in the UK and other Organisation for Economic Co-operation and Development (OECD) countries. Research in other OECD is likely to have less relevance to the UK context and so we will consider the applicability of the international literature to the UK context in analysis, and highlight any limitations on applicability of individual studies. A judgement of the likely relevance to the UK will be made.
We will include any interventions (formal or informal) which were designed to improve, or make better or alternative use of, community infrastructure: physical places and spaces (for example, general urban redesign; interventions such as lighting and benches in open public spaces; children’s play places; or funding to host community activities in places such as libraries or faith settings). We are focusing on interventions that apply at community or neighbourhood level (e.g. a town market), rather than city or national level (e.g. Leeds art gallery). Studies will be excluded if they are not related to a specified intervention, or if they examine a virtual (not physical) space.
We will include quantitative studies which compare different interventions, including those using before and after design and comparing new versus current practice. Qualitative studies without a comparator will be included.
For the purposes of our review, we are defining community infrastructure as:
• Public places and “bumping” places designed for people to meet e.g. streets, squares, parks, play areas, village halls, community centres;
• “Third” places where people meet informally or are used as meeting places in addition to their primary role e.g. cafes, pubs, libraries, shared areas in housing developments, schools, churches;
• Services that can improve access to places to meet e.g. town planning, urban design, landscape architecture and public art, transport, public health organisations, subsidised housing sites, bus routes.
We will focus on interventions operating at the neighbourhood level rather than city or national level, although the focus of the intervention may not be place-based.
We are not including “virtual” spaces such as social media as, although these are important and there is a growing evidence base, we feel that including both real and virtual places (and interactions between the two) in one review would make it too complex and potentially obscure important findings.
We will adopt a broad perspective on the outcomes to be included in the review and will include studies which report any outcome relating to social relations, community wellbeing and related concepts such as social capital and social trust. This includes quantitative (measured), and qualitative (views and perceived) outcomes.
While our primary focus is on outcomes at a community level, we will also include individual level health and wellbeing outcomes, which can be linked to community wellbeing (see Theory of Change, South et al 2017). As many of the desired outcomes would only be evident in the long term, we will also look for proxy measures along proposed pathways to change.
Data extraction, (selection and coding)
Results of the searches of electronic databases will be de-duplicated and uploaded to EPPI-reviewer 4 systematic review management software, which will be used to store information and manage each stage of the review process (Thomas, Brunton & Graziosi 2010).
Studies will be selected for inclusion through two stages, using EPPI-Reviewer review management software. First, a random 20% of all titles and abstracts will be double-screened by all reviewers, followed by a ‘calibration’ exercise to ascertain levels of agreement. Once agreement is reached (80% agreement on include/ exclude), the remaining titles and abstracts will be screened by a single reviewer. Any queries will be resolved by discussion. Full-text copies of potentially relevant studies will be screened for inclusion using the criteria outlined below. Disagreements will be resolved by discussion, with a third reviewer being consulted where necessary. The results of the abstract screening will be recorded in EPPI-Reviewer, while results of the full paper screening will be recorded in EPPI-Reviewer and presented in an Appendix to the review, including the reason for excluding any paper. The study selection process will be presented in a flow chart in the review.
Data from each included study will be extracted into pre-designed and piloted forms on EPPI-Reviewer 4 systematic review management software (Thomas, Brunton & Graziosi 2010). Forms will be completed by one reviewer and checked for accuracy by another. Periodically throughout the process of data extraction, a random selection will be considered independently by 2 people (that is, double assessed) for at least 20% of the studies. Data to be extracted include: study aims, study design, setting/country, intervention, comparator, population, outcomes measured and main findings in relation to the review questions.
We plan to use the Context and Implementation of Complex Interventions (CICI) checklist (Pfadenhauer et al. 2016, page 24) to extract and assess information (where reported) in the following domains of implementation strategy, context and implementation, to assist with answering the review question on process (What factors (positive and negative) affect the implementation or effectiveness of the interventions?), (and see Appendix 4):
• Implementation theory
• Organisation & Structure
Owing to logistical and time constraints, depending on the number of relevant studies located, it may not be possible to contact study authors for any unclear, missing or additional data.
Risk of bias (quality) assessment
We will conduct validity assessment of all studies using the appropriate checklist, following the recommendations of the What Works: Wellbeing methods guide (Snape et al., 2017). Unpublished data from grey literature will be assessed using the same criteria as is used for published data.
Each full paper will be assessed by one reviewer and checked for accuracy by another. Periodically, a random selection will be considered independently by 2 people with at least 20% of the studies being double assessed. Any differences in validity grading will be resolved by discussion or recourse to a third reviewer. Validity assessment data will be extracted and recorded using EPPI-Reviewer review management software.
In this review we propose to be inclusive and use studies that are of ‘low quality’, explicitly describing the implications of including them.
We will examine specific features of the body of evidence, namely type of evidence, validity of evidence, consistency of findings, and consistency between unanswered research questions.
Strategy for data synthesis
Evidence synthesis will use a range of approaches depending on the design of the included studies, including narrative synthesis (Popay et al. 2006), meta-analysis for quantitative studies (Higgins et al. 2008; CRD 2009) if appropriate, and meta-ethnographic approaches for qualitative studies (Noblit and Hare 1988). A mixed method systematic review design similar to that used by the Evidence for Policy and Practice Information and co-ordinating (EPPI) Centre (Thomas and Harden, 2008) will be used to combine data from different study designs. Evidence will be initially synthesised by study type into two streams: quantitative and qualitative (for studies that use mixed methods, qualitative and quantitative data will be extracted and treated separately in the relevant streams).
The narrative synthesis will form the overall reporting framework for the review findings, which will be grouped by review question, setting and by intervention, population or outcome (decisions on this will be data driven with reference to the review advisory group), and will include:
• Thematic analysis of data based on the review questions.
• Exploration of relationships within and between studies.
• Differential impacts in relation to (e.g.) gender, socioeconomic status, ethnicity, or disability status will be considered.
• The strength of evidence will be identified based on study design, and on the results of the critical appraisal (for each type of design).
• Contradictions in findings will be examined.
Preliminary searches suggest that statistical meta-analysis may not be appropriate due to clinical heterogeneity of study designs, outcomes and interventions, but it may still be possible to display quantitative results in Forest plots, without pooling data. If statistical meta-analysis were possible, studies would be combined using a random effects model to give relative risks with 95% CIs for binary outcomes and weighted or standardised mean differences with 95% CIs for continuous outcomes. Statistical heterogeneity would be examined using the chi-square and I-square statistics, with a chi-square p-value of >0.1 or a I-square value of >50% indicating statistical heterogeneity, in which case a random effects model would be used to combine data.
Thematic synthesis using QSR NVivo software will be used to combine evidence from qualitative studies (Thomas and Harden 2008; Oliver et al 2005; Harden et al 2004). This will take place by two reviewers working independently in three stages which may overlap: free line-by-line coding of the findings of included studies; construction of ‘descriptive’ themes; and the development of ‘analytical’ themes (Thomas and Harden 2008). Coded text will be checked for consistency of interpretation between studies and between reviewers. Reviewers will collectively identify similarities and differences between the codes to start to group them into descriptive themes. Analytical themes will then be developed by two reviewers independently applying the review objectives to the descriptive theme (Thomas and Harden 2008).
We will attempt to produce a conceptual pathway of how community wellbeing is related to community infrastructure based on the evidence retrieved.
We will also generate an evidence map, which tabulates the identified evidence in terms of which aspects of community wellbeing and which aspects of community infrastructure (e.g. setting, intervention, population) they address.
Analysis of subgroups or subsets
This systematic review will produce:
• Registration with PROSPERO, an international register of systematic reviews.
• A systematic review
• A conceptual pathway of how community wellbeing is related to community infrastructure, and can be boosted by interventions to improve community infrastructure, based on the evidence retrieved from the reviews
• An accessible summary document
• Journal publications and conference presentations.
Contact details for further information
Centre for Health Promotion Research
Leeds Beckett University
Organisational affiliation of the review
What Works Centre for Wellbeing
Dr Anne-Marie Bagnall, Leeds Beckett University Professor Jane South, Leeds Beckett University Dr Salvatore Di Martino, Leeds Beckett University Dr Kris Southby, Leeds Beckett University Ms Gerlinde Pilkington, Liverpool John Moores University Mr Andrew Pennington, University of Liverpool Professor Rhiannon Corcoran, University of Liverpool
Details of any existing review of the same topic by the same authors
A scoping review of what works to boost social relations was undertaken to identify evidence gaps to be addressed in this review: https://whatworkswellbeing.org/2017/06/20/what-works-to-boost-social-relations/
Anticipated or actual start date
01 June 2017
Anticipated completion date
31 October 2017
What Works Centre for Wellbeing is funded by the ESRC and Public Health England
Conflicts of interest
Subject index terms status
Subject indexing assigned by CRD
Subject index terms
Health Behavior; Humans; Residence Characteristics
Stage of review
Date of registration in PROSPERO
12 July 2017
Date of publication of this revision
12 July 2017
Stage of review at time of this submission
Piloting of the study selection process
Formal screening of search results against eligibility criteria
Risk of bias (quality) assessment
PROSPERO This information has been provided by the named contact for this review. CRD has accepted this information in good faith and registered the review in PROSPERO. CRD bears no responsibility or liability for the content of this registration record, any associated files or external websites.