A public health approach to mobilizing community partners for injury prevention: A scoping review

Objectives Reducing injuries in adults requires work with diverse stakeholders across many sectors and at multiple levels. At the local level, public health professionals need to effectively bring together, facilitate, and support community partners to initiate evidence-based efforts. However, there has been no formal review of the literature to inform how these professionals can best create action among community partners to address injuries in adults. Thus, this scoping review aims to identify theories, models or frameworks that are applicable to a community-based approach to injury prevention. Methods Searches of scientific and less formal literature identified 13,756 relevant items published in the English language between 2000 and 2016 in North America, Europe and Australia. After screening and review, 10 publications were included that (1) identified a theory, framework or model related to mobilizing partners; and (2) referred to community-based adult injury prevention Results Findings show that use of theories, frameworks and models in community-based injury prevention programs is rare and often undocumented. One theory and various conceptual models and frameworks exist for mobilizing partners to jointly prevent injuries; however, there are few evaluations of the processes to create community action. Conclusions Successful community-based injury prevention must build on what is already understood about creating partnership action. Evaluating local public health professional injury prevention practice based on available theories, models and frameworks will identify successes and challenges to inform process improvements. We propose a logic model to more specifically guide and evaluate how public health can work locally with community partners.

intersectoral partnerships within health promotion described positive processes for improving population health more generally [15]. How diverse community partners become mobilized to achieve reductions in unintentional injuries remains unclear [9]. While listing promising programs, Ontario's "Prevention of Injury Guidance Document" currently presents no theories, frameworks or models for how local public health professionals can mobilize community partners in effective injury prevention [13]. Proven interventions must have adequate support to be effective, thus, local public health professionals must also understand how to facilitate community-wide partnerships to produce socially significant injury reductions [16].
Thus, the purpose of this paper was to conduct a scoping review to identify theories, frameworks, and models (TFM) that clarify how to mobilize partners in community-based adult injury prevention and to what effect, seeking evidence from a range of real-world sources rather than being limited to any particular quality standard (e.g. randomized controlled trials). Given there are no published literature reviews, a scoping review is a systematic way to scan and aggregate existing literature to further articulate the concept of mobilizing partners, identify knowledge gaps and inform future directions related to community-based adult injury prevention.

Project team
This project involved a multidisciplinary team including nine public health professionals from Ontario public health units, two academics, an academic librarian, a graduate student, and a research consultant, with collective expertise in injury prevention, scoping reviews and knowledge translation activities. Member engagement varied, with two project leads managing the collaboration, the consultant coordinating the research project, 6 team members reviewing literature, and knowledge users providing stakeholder advice. All members participated in biweekly teleconferences, building relationships to integrate knowledge translation between researchers and practitioners. In addition, three face-to-face meetings actively involved endusers at the various stages of the research, with each contributing their own distinct expertise and experience to the process.

Identification of relevant studies
A methodological framework, developed by prior authors, guided our scoping review process [17,18]. The first step entailed defining the working terms (see Table 1).
Our literature search included documents of any study design and publication type. Articles were selected based on the following inclusion criteria: 1) published between 2000 and 2016; 2) published in the English language; 3) related to community-based injury prevention among community dwelling adults; 4) presenting one of a TFM for mobilizing community partners; and 5) based on literature in developed countries similar to Canada such as the United States, Australia, New Zealand and high income European nations. We excluded articles that were 1) book reviews, editorials or commentaries; 2) not in a developed high income country; 3) without any TFM for mobilizing community partners; 4) not related to community-based injury prevention; and 5) not addressing an adult population of 18 years and older.
The librarian member developed search strategies to identify relevant articles in consultation with Public Health Ontario researchers and Shared Library Services. Subject headings and keywords included: 1) theory/framework/model (e.g., paradigm, strategy, approach); 2) mobilization (e.g., capacity building, coalition, collaboration, community network, multi-sector partnership); and 3) injury prevention (e.g., accident prevention, falls, protective devices, safety). . Hand searches were also performed using 1) the reference lists of the included items; 2) the tables of contents of three journals that publish articles related to injury prevention and public health (e.g. Health Promotion Practice, Health Promotion Journal of Australia, American Journal of Public Health); 3) additional suggestions made by project team members; and 4) publications of eight well-known authors in the field of community health partnerships identified by our academics (i.e., Neil Bracht, Francis Butterfoss, Stephen Fawcett, Paul Florin, Michelle Kegler, Meredith Minkler, Nina Wallerstein, and Abraham Wandersman).

Study selection
As shown in Fig 1, the initial search yielded 13,756 sources of literature. A database management system (EndNote) removed 2371 duplicates. We established a reviewer group that included the research consultant, an academic, a graduate student, and three public health nurses. We calibrated consistency in determining study inclusion and exclusion was among all reviewers, first through independent assessments (yes, no, unsure) of 25 random abstracts followed by paired assessments of 75 random abstracts. Reviewers agreed on 83% of the abstracts to be included. We discussed and agreed upon differences via consensus. Following these

Term Definition
Theories Systematic views of interrelated concepts, definitions, and propositions to explain and predict events or situations [19].

Frameworks
Broad structures of descriptive categories containing both explicit and assumed propositions [20].

Models
Symbolic representations of theoretical concepts for further understanding of a problem in a specific context [21,22].

Mobilization
Moving a group of people into collective action.

Community Partners
Organizations with a vested interest in injury prevention or an interest in the well-being of adults living independently with some common characteristic, for example locale, faith or heritage. Living independently includes in group homes and the homeless but not those in institutions such as hospitals, treatment facilities, long term care homes, jails, or similar. For our purposes community partners may include governmental agencies, non-governmental agencies, coalitions, networks, industry, employers, location-specific organizations, health authorities, healthcare institutions, etc. Citizens were not included as partners to be mobilized but rather as clients of public health services and drivers of any improvements to their health and wellbeing.

Community-based Adult Injury Prevention
Reduction of the frequency, severity and impact of injury among adults (18+) through community involvement. The community-based injury prevention model emphasises that community participation and multidisciplinary collaboration are needed to address local problems [23].
https://doi.org/10.1371/journal.pone.0210734.t001 exercises, we distributed all titles and abstracts among the reviewer group to screen for inclusion by asking two questions: "Does the title or abstract identify a theory, framework or model related to mobilizing community partners?" and "Does it refer to community-based adult injury prevention?" Reviewers excluded all but 251 items and these advanced for a more indepth review of their full text. We then distributed full-text articles among the reviewer group to determine further exclusions using the screening criteria as well as an additional criteria "Is the context of development and/or implementation of models, theories and frameworks in a developed country similar to Canada (e.g. United States, Australia, New Zealand, high income European nations)", which may not have been apparent from the title or abstract alone. Any full text articles rated as "unsure" by the first reviewer were assigned to a second reviewer and discussion determined final eligibility for inclusion. Reviewers presented the 10 publications that satisfied our criteria for inclusion to the full project team to verify perceptions of relevance to our research question and applicability within public health practice, as well as for knowledge exchange.

Data charting, collating, summarizing, and reporting
A data extraction tool captured the reference source, literature type, study design, and the identified TFM and incorporated a critical appraisal tool for the included literature. The Meta-QAT or the Public Health Ontario Meta-tool for Quality Appraisal of Public Health Evidence [24] is a novel critical assessment tool developed to examine article quality (i.e. relevancy to our research question, applicability to public health unit practices in Ontario, reliability of design and validity of results). Using one of the included articles, project team members (consultant, student and public health nurse) independently tested the data extraction tool for clarity and ease of use. The MetaQAT does not provide a numerical score for rating articles; rather, long-form responses recorded the identified strengths and weakness of each literature source. At least two project team members appraised and summarized the ten included articles.
As not all included articles provided complete descriptions of the identified TFM, supplementary sources provided the full list of key elements comprising each TFM (e.g., constructs, steps, components, phases, tasks). Two project team members (CD, SL) pragmatically sorted the elements of each TFM into three broad categories related to 1) prerequisite conditions for community partner activity (e.g. formation as a group), 2) results of community partner activity (e.g. achievements), and 3) neither prerequisite or resultant, thereby distinguishing transitional elements to conceptualize partner mobilization. Subsequently, the full multidisciplinary team reviewed, refined and confirmed the associated categories.

Results
The ten sources collectively identified one theory, seven frameworks, and six models for mobilizing partners. Three publications identified more than one framework or model [11,25,26]. Five of the sources were specifically set in an injury prevention context [23,[26][27][28][29] with two focussed on generally preventing injury across all ages [23,26], two focussed on preventing recreational and professional sport injury among adult men aged 18 to 35 years old [27,28] and one focussed on preventing fall-related injury among adults 65 years of age and older [29]. The remaining five sources addressed adult injury prevention more generally as part of population health improvement initiatives [11,25,[30][31][32]. Five sources provided brief descriptions of a TFM [11,25,28,30,32], four sources proposed a new framework or model [23,26,29,31] and three described applications [25,27,28]. Table 2 summarizes the included sources based on the relevance to the research question, applicability to local public health agencies, reliability (or reproducibility) and validity (rigor of research). Four of the ten sources were primary articles including two case studies and two were quasi-experimental (repeated measures) design. One case study applied Step 5 of the Intervention Mapping Framework [33] to prevent Australian football injuries in adult males [27]. The second case study presented 12 core components across four stages of coalition building to propose a Coalition Development Model for successful collaboration [26]. One study used a repeated measures design to evaluate a multi-agency partnership between the government (health promotion division), safety agencies, sports professionals, advocacy bodies and health insurance organizations to reduce recreational sports injuries over a 5-year period [28]. Another study used a repeated measures design to monitor physical activity programming intended to reduce falls and related injuries among older adults, reporting a process and impact evaluation of the Collaborative Management model that supported a promotional campaign [29].

Characteristics of the included literature
Five of the ten sources were secondary literature: four reviews (three narrative reviews and one scoping review) and one synthesis of evidence for community action. One author presented 5 components and 12 processes to guide communities' capacity to improve their health based on the Collaborative Public Health Action Framework [30]. Another author presented a typology and factors influencing effectiveness of multi-sector partnerships to improve population health [11]. The third author presented an evidence-informed logic model for community-based injury prevention programming [23]. The scoping review author synthesized 14 critical steps for implementing quality improvement innovations [31]. The synthesis of evidence for community action was the Centre for Disease Control's Sustainability Planning Guide for Healthy Communities designed to help coalitions, public health professionals and community stakeholders develop, implement and evaluate sustainability plans. [25].
One of the ten sources was tertiary: an online brochure for community health improvement using the model of "Mobilizing for Action through Planning and Partnerships" [32]. Table 3 illustrates the transition from establishing prerequisites to achieving results based on the key elements of the TFMs identified by the included literature. Table 4 shows the common steps and elements within each theory, framework and model for developing sustainable and working partnerships. The common elements were grouped based on the timing of the partnership process and distinguish prerequisite and resulting conditions from the transitional, and hence "mobilizing", elements. Prerequisite conditions to mobilizing consist of partnership start-up, community assessment and strategic planning; resulting conditions are centered on evaluation; and ongoing involvement, instrumental supports and planned action describe transitional mobilizing elements.

Theory, frameworks and models partnership activities
Pre-requisites to mobilization. Partnership Start-up. All but one TFM noted the importance of developing partnerships. However, the literature had different accounts of what aspects were critical to developing partnerships. United in commitment was most frequently mentioned (86%), followed by structured meetings (71%), recruitment of stakeholders and key organizational members and partners (57%) and skill building activities (43%).
Community Assessment. Twelve of the TFM noted the importance of performing community assessment, defined as determining priority areas for intervention, as part of building productive partnerships. Most noted understanding the issue or problem, and the sharing of  resources and active participation (planning, meetings, skill building), 64% and 71% respectively, in all the TFM's. Strategic Planning. Strategic planning consists of determining how an intervention developed through partnership work will result in improved outcomes. While 64% found that planning and adaptation (developing strategic plans, action plans, interventions, policies) was an important pre-mobilization activity, only 21% of the TFMs mentioned the need for a logic model to guide programs and services.
Transition to mobilization. Ongoing involvement. More than half (57%) of the TFM identified ongoing involvement with community members, partnerships and stakeholders, categorized as participation (43%), commitment (29%), leadership (21%), and collective aim (14%). One theory, three frameworks, and two models identified member participation. One theory, two frameworks, and one model identified member commitment (29%) as engagement, empowerment, and high quality, trusting relationships.
Instrumental supports. More than half (57%) of the TFMs identified instrumental supports, categorized as collective resources (57%) and public support (21%). One theory, five frameworks, two models identified collective resources, including material resources such as data and sustained funding, human resources such as a designated implementation team, and relational resources such as additional partners outside the partnership that can help with mobilization efforts with the greater community. These resources among core members or across the broader partnership should be shared for mutual benefit and common purpose. One theory, one framework, and one model identified public accountability, including context and community support that could be generated generically through grassroots development and community organizing or specifically through messaging, networking, reaching out and staying visible in the media.
Planned Action. More than half (64%) of the TFMs identified planned action, categorized as planning (57%) and implementation (43%). One theory, four frameworks, three models identified planning, as designing outputs such as effective policies, services and environments, designing processes such as strategies for change, communication, and social marketing, and designing actions such as implementation and program operation. One theory, two frameworks, three models identified implementation related to what was designed: effective policies, services and environments, strategies, actions, a sustainability plan, program operation (e.g. implementation).
Mobilization Results. Evaluation. Post-mobilization activities consist of evaluating to determine whether an intervention, program or service has had positive effects on the population of interest. Only 50% of the TFMs identified evaluation activities post mobilization including carrying out evaluations plans (36%), evaluating the implementation of a program or service (21%) and sustaining service delivery (21%).

Discussion
Although none of the TFM includes all the components necessary for forming and sustaining productive partnerships [23], the findings show that there are common features among the TFMs. These common features suggest that there are preconditions and transitional elements needed to achieve results, thereby contributing to the overall understanding of what successful Collaborative Management [29] N/A Collaboration between state and local levels were not found to be a suitable approach for the project's purpose N/A Partnerships for health improvement [11,40] • To extend the reach and capacity of governmental public health who has primary responsibility • To deliver public health services with others who play some role in promoting public health.
• To address public health through all sectors of community life (eg, education, business) as the system of actors and actions that promote or threaten population health • To create change primarily by impacting the context through the relationships among organizations in the partnership Note: Each bullet represents an element identified in the theory, framework or model for establishing and maintaining coalitions and partnerships.
https://doi.org/10.1371/journal.pone.0210734.t003   mobilization should encompass. While each of the TFMs offer conceptual insights into partner mobilization, they independently lack detailed evidence needed to inform, support and guide public health practitioners who are mandated to advance community-based adult injury prevention. Our findings also suggest there is less attention dedicated to maintaining and evaluating community-based adult injury prevention work, particularly in the mobilization and post-mobilization phases. We found only one process or outcome evaluation among the studies reviewed indicating a lack of evidence to support investment in such partnership work similar to a prior review that found that behavioural and social science theories and models are seldom represented within injury prevention research [41]. Although the logic model by Nilsen [23] provides theoretical considerations for what works, how and why, very few studies we reviewed included an assessment of reliability and validity. More than half of the reviewed documents did not provide information on the methodologies used to construct their respective TFM. Therefore, when considering all of the TFMs in this review that included the need for evaluation, we developed a logic model that outlines the required inputs, outputs and desired outcomes related to injury prevention initiatives in community settings (see Fig 2) to articulate what a partnership effort needs, what it expects to do, how it will work and what success it hopes to achieve.
The prerequisite conditions for community-based injury prevention initiatives form the logic model inputs, and include partnership set-up, community assessment, and evidencebased strategic planning. Involvement from the outset can result in trust building, membership, and ownership among partners for better program implementation [42]. Of the TFMs found in our review, only 57% included ongoing involvement suggesting that many current TFMs perhaps do not reflect the importance of involving stakeholders and organizations throughout program planning and implementation as central in a community organization approach [43].
Partner mobilization forms the logic model outputs, and includes ongoing involvement, instrumental supports, and a detailed course of action, within an ecological framework of the community organizations and networks that can contribute to the context for injury and its prevention among community-dwelling adults. Application of the ecological model in injury prevention holds the most promise in regard to fall and traffic injury prevention, as well as community safety promotion programs [44]. Results form the logic model outcomes, and include building awareness and knowledge of evidence-based interventions, creating enabling environments, and effective community action to achieve injury reductions. A health promotion approach is particularly useful for injury prevention because it specifically facilitates both behavioural and environmental change [45].
Our logic model is consistent with a recent literature review of the components for developing and reinforcing positive partnerships. Partnership start-up includes shared mission, trust and shared power, clarity of roles and structure and appropriate communication. Sustained active involvement includes collaborative leadership. Instrumental supports include active contribution of financial and human resources and consideration of context. A prepared course of action balances maintenance and production. The logic framework format supports evaluation of outputs and outcomes [15].
Mobilizing partners requires attention to sustaining their involvement, obtaining the resources needed (including time and knowledge), and having a clear course of action acceptable to the adult community. In an evaluation of the Healthy Communities movement, findings show that initiatives often rely on community volunteerism and lack the well-supported infrastructure that is essential to successfully maintain ongoing partner involvement [46]. Instrumental supports may come directly from partnering organizations or from the connections and relationships each partner has within the community (i.e. community networks), thereby establishing an enabling context by providing the resources and visibility needed. Such accountability is foundational to mutual trust between public health and communities for building the community's capacity to address injuries, consistent with findings from a prior review [15]. Communicating and delivering detailed actions keeps implementation transparent and timely, as instrumental supports (i.e. funding, time, accountability, using data to inform policy and practice) may otherwise subside in practice. Thus, it can be seen that collaborative participation in making change, so valued in public health, will lead to program implementation, evaluation and knowledge translation activities that will result in a reduction of injuries.
While it is clear that evaluation-based activities are needed to assess short and long-term outcomes of partnership development, measuring the success of both partnership formation and project management is essential to injury prevention initiatives. This model provides a framework for fully measuring the success of building productive partnerships with definitive outcomes related to injury prevention. Evidence from measuring the success of community-based partnership development and resulting action is required to justify further public health expenditures. Without literature evaluating community-based adult injury prevention through partnership, public health practitioners will continue to struggle to mobilize and evaluate partnership work with reduced resources. Under these conditions, it is not surprising that intersectoral collaboration is a challenge to maintain, particularly as other non-profit organizations, businesses and health care providers also need evidence to justify investment as a community partner. Now more than ever, community partnerships and local based programs are needed given the high financial strain on the health care system from unnecessary injury related hospital visits. Future work in this field requires a greater emphasis on measuring short and long-term outcomes related to community-based injury prevention initiatives, as well as public health efforts to collaborate with multi-sector partners, to enhance both the reliability and validity of such mobilization and programmatic efforts.

Study limitations
Although our team members followed a collaboratively developed and refined the search protocol (inclusion and exclusion criteria), it is possible that we missed some sources of literature, particularly as we searched through multiple large bodies of sources. Additionally, we may have missed sources of literature published prior to 2000. We also did not provide evidence ratings of each included article, however, we did provide qualitative feedback using the Meta-QAT. This alone should help in assessing reliability and validity, and evidence ratings could be a valued contribution when a more guided systematic review is performed in the future.

Conclusion
In summary, this scoping review synopsized the theories, frameworks and models (TFM) related to community-based injury prevention found in the available literature. We propose a logic model that combines important components from all the included TFMs into a promising process that better underscores the inputs and activities needed for public health practitioners to build and sustain partnerships active in preventing injuries, as well as plan for evaluating their success.