A study protocol for applying the co-creating knowledge translation framework to a population health study

Population health is concerned with health outcomes within an aggregated group of people, and research on population health considers the impact of environmental and system variables on the distribution of those health outcomes [1]. It is a research domain that has received limited attention within the KT field. Two aspects have been identified as being relatively absent in population health studies: non-quantitative studies and the incorporation of people as agencies within a health system [2]. This protocol describes how the co-KT Framework incorporates KT into a population health research study.

The impact of KT on population health studies is of interest to the research community, as there is a need for empirical studies of the effectiveness of KT in improving the health of the population by changing practices [3]. Others have identified priorities for KT development work such as the need for better structural conceptualising of KT and the need to adapt KT strategies to context [3]. Further it has identified that the majority of researcher-practice frameworks address only one phase (sometimes two) of the process of research-practice translation [4].Within this study, we apply the co-KT Framework to that phase of KT concerned with moving discoveries into communities, and then extended to organisations and policy developers, which has been described as T3 [5]. This area is still being defined in scope [6].

Population health has been the subject of limited KT research approached in a systematic way [7], and there are very few KT intervention studies with a focus on public health and prevention and whole systems redesign [8].Lightfoot et al. (2008) describe ways of conducting population based health research in rural areas, emphasising early stakeholder involvement if the study is proposed by external researchers and suggesting a community needs assessment to help to build community and funding support for the proposed study [9]. Ogilvie et al. discuss a translational framework for public health research, shifting the focus to the improvement of population health through individual and collective determinants of health [10]. This is a clear difference from embedding evidence-based interventions. In public health planning frameworks generally, tacit knowledge is not referred to as a legitimate source of knowledge nor do they pragmatically suggest how this knowledge might be elicited [11].

This protocol describes how KT is being conducted within The Physiology of Health Systems: Port Lincoln as a Case Study (the LINKIN Health Study, National Health and Research Council project grant 627240, 2010–12). The research team is finding a way to link evidence to issues within a particular context rather than applying evidence and identifying barriers. The LINKIN Health Study is a mixed methods study of the population of Port Lincoln, South Australia (population: 14,000 people) [12]. The approach described in this paper is intended to provide detailed insight into the use of both formal and informal sources of health-care and into the extent and pattern of any mis-alignment between established morbidity and use or provision of services. This is in contrast to solely using epidemiological tools to map person-based data on use of health services and limiting the evidence base to information collected by quantitative methods.

There is no favoured operational KT approach for a total research project process. We found that KT frameworks tended to separate the knowledge generation process from the actual implementation of knowledge [13‐15]such as those that addressed evidence-based practice or the diffusion of knowledge process, the effectiveness of KT implementation, in the form of interventions, or the uptake of evidence and articulating conditions for successful KT [16‐22]. The KT framework we have applied (the co-KT Framework) does not begin by applying evidence-based practice but concentrates on allowing the emergence of context specific data.

Our framework is presented as a means whereby KT within a research study may provide a structured means of developing a shared understanding of a problem and its underlying factors [23].

Recognition of the move to develop community interventions that take account of the complex array of causal factors underlying health inequalities has been important to the study team [24]. We also wanted to shape a methodological approach that was not confined to a single-issue intervention but rather used the research process to open up the possibility of addressing community systemic conditions [24]. The framework goes some way to working on the little addressed aspect of the ways in which the interaction between multiple players in a system might be managed and knowledge intermediation (managed processes of knowledge interaction) [25, 26].

This protocol describes how a theory informed conceptual framework for KT will be applied in a population health study that sets out to formalise collaborative approaches within the design, whilst differentiating between the roles of both the stakeholders and the researchers. The protocol describes the application of a KT framework to a mixed methods population health study. A separate protocol relates to the mixed methods design itself. The protocol demonstrates the deliberate incorporation of the theories of engaged scholarship [27, 28] and practice of participatory action [29] as we put research generated knowledge as a key driver of improved health care. In particular, we see these underlying theoretical foundations as enabling the co-learning and co-construction of knowledge and understanding of that knowledge, the acknowledgement of a range of social and environmental factors that impact on health, and including community involvement at all stages of the research process [29]. Action research and engaged scholarship explicitly promote the concerns of the local context and have the capacity to generate knowledge that addresses local concerns [30]. There is an assumption within these two theories that there may be multiple sources of knowledge and that people within the community possess valid knowledge. We do not suggest knowledge is objective nor a set of context-free facts that must be packaged for communities but rather we are responding to the practical recognition that knowledge translation and intermediation can be sites of conflict [25, 26].We view knowledge as a research tool and point of common engagement; knowledge is to be constructed and shaped, in our study, for the purpose of improved population health.

Action research and engaged scholarship both focus on the researcher– community relationship as a combination that will yield effective research rewards [30]. We anticipated that this theoretical base would facilitate the framing of research questions that were more meaningful to the community (for example, rather than specifying a reduction of ill-advised rates of health incidence, we could frame our issues for address in terms of what might most benefit the community). There has been recognition of the interface between participatory oriented research and KT and the co-construction of knowledge in a relationship that includes stakeholders and researchers [31]. Community-based participatory research methods have directly addressed the researcher and community relationship, and its role in the community intervention process [24]. Clavier et al. found that strategic translation supports the research process and facilitates the ongoing collaborative involvement of stakeholders [31]. A further point of congruence between KT and participatory action research is that community-based participatory research places significance on actionable knowledge, and knowledge produced through relationships and collaborative practice [29]. It also considers the problem of synthesising evidence which is important in a mixed methods design.

The LINKIN Health Study examines the health system functioning within the rural population of a South Australian town and takes place at the nexus of university and community research. It aims to build knowledge for the purpose of health reform between these two contexts, using quantitative and qualitative data collection as a foundation on which to develop suitable health interventions [12]. The co-KT Framework is to form the integrating and synthesising thread through the research process and is led within the research team by a nominated sub-group which is a key feature of the study protocol.

The co-KT Framework is a theory based model for knowledge translation practice that has been shaped by the requirements of a particular population health study of a mixed methods design. We see the co-KT Framework as meeting a need faced by researchers, and in particular, multi-disciplinary teams of researchers when they strive to conduct a community-based study or population health study, to achieve research aims through the development of context- relevant knowledge. This resonates with the call from Hartling et al. (2007) who state that there is an urgent need for a better understanding of KT interventions from a multi-disciplinary perspective, KT interventions that are designed to target multiple professional groups, and that health care delivery improvements will come from interdisciplinary collaboration [44]. There is also a growing movement to reduce the distance between knowledge development and knowledge use, and linking this with the need to develop community interventions that take account of several sources of health inequalities [24]. To address such complexity, there must be opportunity to have an understanding of how each party to the research process (researchers, communities and stakeholders) frame a problem [45]. We believe we have opportunity for discursive discussion within the framework steps.

The co-KT Framework recognises the translation derived elements separated by Clavier et al. needed to operationalize a framework; incorporating cognitive practices (managing the research contents), strategic methods (to conduct the research process and manage relationships across stakeholders) and logistic practices (that relate to practical coordination) [31].

KT has been criticised for not considering sufficiently methods whereby the end-users of interventions or innovations might critically reflect on contextual related factors that might influence the embedding of an innovation within a context of practice [46].This protocol activates the engaged scholarship theoretical basis of the co-KT Framework in its participative form of research for obtaining the perspectives of different stakeholders and for recognising different sources of data to develop the most useful knowledge to address complex problems. Action research and PAR and engaged scholarship contain principles of enjoining and involving both academic researchers and situationally defined stakeholders within the dynamics of building knowledge around a framed issue. The protocol in particular illustrates how the expertise and knowledge of researchers and communities might be brought to bear on the health systems of a regional place [28]. The co-KT Framework incorporates the study context into the process of knowledge to evidence and evidence to innovation design, inviting critical reflection. Here, we demonstrate in this approach a view that community engagement for KT is more than smoothing a path to the implementation of interventions; it allows for an understanding of how public health policies and local context factors combine structuring the system, the reality, that must be navigated by individuals as the end-users to self-manage their health.

We appreciate that in community-based KT work there can be tension between what is considered evidence between clinicians and community-based organisations [47]. Kothari and Armstrong state that traditional KT approaches may not be suitable for all health service delivery systems [47]. Co-KT, as an iterative knowledge creation and translation process, may be better suited to community-based research. By incorporating KT at the knowledge creation (evidence building) stage, we have emphasised our focus on health system re-design and treatment of all stakeholders as part of a total health system serving a defined population.

An important aspect of the incorporation of KT within a population health study we suggest is the direction of community engagement into a form of social action that results in a health system change for the community that goes beyond simply research participation or engagement by a university [2]. How well this will be done remains to be seen but it is important at this stage of the protocol design that inclusion of the community in the knowledge building is more purposeful than local relationship building, and allows for the articulation by stakeholders of change options.

The protocol recognises the value of the social processes used to undertake our research [2]. We have emphasised face to face communication as a key form of relationship building and communication within our framework [29] and recognise that this poses its own dynamics and difficulties with organisation. However, we also consider this a better way to influence the credibility of the outcomes of our research [29].

We are aware that there is a shortage of rigorous operational studies or evaluation of KT strategies [8]. Dagenais et al. (2009) note that few studies present results on the effects of populations that are the subjects of interventions [3]. The co-KT Framework incorporates methods of evaluation within each step and at the completion of the research study and attention will be given to collating evaluation data from the LINKIN Health Study. We view the consulting itself as an interactive knowledge transfer tool and its effectiveness will be targeted for evaluation by qualitative means [48]. Methods of evaluation include feedback questionnaires, non-participant observation, interviews, assessment of the relevance of documentation, and intervention specific data. Simple knowledge and research utilisation questions will be employed as these have been demonstrated to be effective [49]. One of the dimensions intended for evaluation is ‘population reach’, that is, how many people did interventions resulting from the research impact upon?