Article Text

Integrated knowledge translation: digging deeper, moving forward
  1. Anita Kothari1,
  2. C Nadine Wathen2
  1. 1School of Health Studies, The University of Western Ontario, London, Ontario, Canada
  2. 2Faculty of Information & Media Studies, Centre for Research & Education on Violence Against Women & Children, The University of Western Ontario, London, Ontario, Canada
  1. Correspondence to Dr Anita Kothari, School of Health Studies, The University of Western Ontario, Labatt Health Sciences Building, Room 222, London, Ontario, Canada N6A 5B9; akothari{at}


Background Integrated knowledge translation has risen in popularity as a solution to the underuse of research in policy and practice settings. It engages knowledge users—policymakers, practitioners, patients/consumers or their advocates, and members of the wider public—in mutually beneficial research that can involve the joint development of research questions, data collection, analysis and dissemination of findings. Knowledge that is co-produced has a better chance of being implemented.

Discussion The purpose of this paper is to update developments in the field of integrated knowledge translation through a deeper analysis of the approach in practice-oriented and policy-oriented health research. We present collaborative models that fall outside the scope of integrated knowledge translation, but then explore consensus-based approaches and networks as alternate sites of knowledge co-production. We discuss the need to advance the field through the development, or use, of data collection and interpretation tools that creatively engage knowledge users in the research process. Most importantly, conceptually relevant outcomes need to be identified, including ones that focus on team transformation through the co-production of knowledge.

Conclusions We explore some of these challenges and benefits in detail to help researchers understand what integrated knowledge translation means, and whether the approach's potential added value is worth the investment of time, energy and other resources.

Statistics from


In early 2013, we addressed the question ‘Should practitioners, administrators and policymakers become involved in research?’, that is, engage in a partnership approach to health research that has come to be known as ‘integrated knowledge translation’ (IKT).1 That paper identified some fundamental assumptions underlying the IKT process that are not always made explicit. For one, it is assumed that resources (staff, money) are available for partnership building and development which takes time and effort. Another assumption is that researchers and knowledge users understand and accept that efforts that go into IKT partnerships are not traditionally acknowledged or valued in what today is a very outcome-oriented environment. Third, it is taken for granted that researchers should bear the responsibility of initiating and maintaining IKT partnerships with knowledge users. Finally, there is the expectation that the IKT partnership will result in a definitive outcome that can transform policy and practice, which can be termed a positivity bias. The reality, however, is that often projects can result in a small, possibly inconclusive finding that contributes to a larger body of research but perhaps not to immediate action. These assumptions could potentially lead to disappointment in an IKT partnership if not understood and accepted from the start. However, we pointed out that IKT can also offer a wide range of benefits for those involved. IKT can be a transformative process in which a new group identity is developed, which is characterised by a shared language, world view and understanding of the surrounding context. This process of creating a shared perspective goes beyond the original goals of IKT—to enable better uptake of research and knowledge, once created—and instead focuses on researchers and knowledge users learning to truly understand and appreciate each other's points of view, and to converge on a newly created shared perspective.

Additionally, in jointly generating knowledge, IKT shakes free of the traditions of academic knowledge creation and allows for other actors (practitioners, policymakers, advocates or the general public) to enter the research space as experts in their own right. Finally, the paper argued that something intangible is built among the group during the IKT process—this can be interpreted as tacit knowledge, social capital or relational capital. These intangibles can forge and sustain strong relationships and drive future collaborations, even in the absence of a concrete project-specific outcome. We concluded that even if an IKT project does not meet initial expectations of actionable outcomes and uses significant time and resources, it can still have subtle but powerful benefits that reach beyond the specific initial project.

A 2016 scoping review of IKT studies reported that lack of skill or understanding of IKT processes was a barrier to the process.2 This paper, grounded in the literature, our research and experience, updates developments in the field of IKT, and offers a deeper analysis of the IKT approach in practice and policy-oriented health research. The purpose is to challenge how we currently conceptualise IKT and to examine, briefly, tools used to engage knowledge users. The focus is on the researcher perspective, and specifically for those who want to engage more fully with the approach. The paper is structured as follows. We first discuss various collaborative working models that do not represent IKT approaches. Then we describe some co-production traditions that have informed our understanding of IKT, point to demonstrated gaps in the empirical literature and note the underdevelopment of IKT theory. The section that follows pushes the field by highlighting newer IKT configurations and methods of knowledge user engagement. Before concluding, we raise concerns related to IKT risks.


IKT: what it is not

To begin, we outline our own position on what IKT is not. IKT is not what is sometimes called team science2 ,3 for which a multidisciplinary group of researchers, sometimes across a large geographical area, is brought together to partner for the purposes of sharing research resources, training or expertise. IKT differs from team science in that its core aim is to develop research questions that are informed by needs in practice or policy. IKT is also not a practice-based knowledge sharing or quality improvement network (eg, Stop HIV/AIDS4) where the primary aim is to disseminate research to practitioners rather than generate research and knowledge collaboratively.

There is less certainty about whether action research, community-based participatory research (CBPR), or their variants are, or should be, classified as IKT. The Canadian Institutes of Health Research, a funder which has had a major influence on the worldwide development of knowledge translation research and practice, suggests that CBPR is indeed a type of knowledge translation methodology.5 From our own readings and experiences, we would argue that IKT is somewhat different from CBPR in that the invited IKT knowledge users usually have the authority to invoke change in the practice or policy setting—this is not often, or explicitly, the case in CBPR. Another difference between IKT and CBPR is that the former is explicitly concerned with identifying practice/policy changes or interventions that are based on research findings.

Finally, implementation science (also called translation research) is an approach that overlaps with IKT in that it endeavours to incorporate, or ‘scale-up’, an existing programme or treatment with demonstrated efficacy into practice sites.6 ,7 In these types of projects, partnerships with knowledge users are often developed to understand local barriers and facilitators, and researchers are interested in exploring issues related to fidelity, adaptation and outcomes.

IKT: what it is

The concept of knowledge co-creation is taken up by a number of contemporary scholars. Gibbons et al8 discuss at length the need to revisit the social contract between science and society. They position traditional research processes (called mode 1 research) as unidisciplinary, researcher-driven, subject to peer review and based on a positivist epistemology. In contrast, they describe the new mode 2 knowledge-generation processes as multidisciplinary, inclusive of stakeholder participation and needs, and working from a problem-solving epistemology. From a management science perspective, Van de Ven and Johnson9 argue that the area of knowledge translation arises from the view that research is directly applicable to practice, and therefore a focus on improving transfer strategies ought to improve practice and policy. They suggest, however, that knowledge of theory is distinct from knowledge of practice. Van de Ven and Johnson's9 idea of engaged scholarship—where researchers and practitioners co-produce knowledge—is a reconciliation of these perspectives, and is a way for practice-relevant research to be generated. Both mode 2 research and engaged scholarship align with the IKT model of knowledge production presented above; a robust conceptual analysis of a range of collaborative research traditions is currently underway. The essential elements underlying these frameworks are that: researchers and practitioners/policymakers come together to work on a complex problem relevant to the real world; there is a recognition that both communities hold distinct norms and values but they also bring valuable knowledge to the problem; and the work of knowledge generation is done collaboratively.

Much work is required to support the study and application of these traditional IKT models. For example, empirical work can uncover the outcomes and impacts related to collaborative knowledge production.10 A 2016 scoping review of 13 IKT studies2 found that it was difficult to identify beneficial outcomes across studies given the limited reporting and/or evaluation of IKT activities. Nevertheless, seven studies identified positive outcomes related to researcher and knowledge user capacity development. Among the 13 articles, outcomes seemed to be independent of the initiator of the collaboration, dedicated funding, collaboration maturity, nature of knowledge user involvement, presence of particular, or more, enablers than barriers, or the number of different IKT activities. Surprisingly, common barriers and enablers did not emerge across the majority of these studies, although four of them noted the advantage of multiple and varied opportunities for interaction and five studies suggested that differing needs and priorities were a barrier. A 2014 scoping review of 19 IKT studies focused on rehabilitation research also found lack of evidence related to outcomes, but perceived outcomes included the formation of collaborations, support for the research and its implementation, and the empowerment of knowledge users.11 Both scoping reviews recommend that researchers report their IKT strategies in detail, including how knowledge users were engaged (eg, team meetings, working groups, web portals and training sessions), at which points during the research process users were involved (eg, question development, data interpretation, dissemination) and the intensity of that involvement or level of engagement. A body of research is needed to identify promising approaches to supporting IKT in different organisational settings. The development of outcomes to measure the effectiveness of IKT models is also required.

Finally, theoretical developments supporting IKT approaches are also in short supply.10 Kitson et al12 contribute to this gap in knowledge by developing a co-KT Framework based on a critical review of the literature and their own case study. The authors describe co-KT as “a framework for actioning the intent of researchers and communities to co-create, refine, implement and evaluate the impact of new knowledge that is sensitive to the context (values, norms, and tacit knowledge) where it is generated and used” (p. 3).12 This Framework encompasses five steps through which information is shared between the researcher context and the study context. The researcher context represents the site of research knowledge generation, methodological attention and the facilitation of exchange processes. The study context is the site where improvements are meant to happen, and where context-specific information is generated. Issues are framed; research and local knowledge are synthesised, adapted and tested; and interventions are evaluated and embedded through the five-step framework. Essentially, the authors acknowledge the need for universal and local knowledge to direct change efforts. Additional theorising like this preliminary Framework can result in a stronger foundation for the IKT field.

IKT: what it could be

New IKT models

The fundamental concept underlying IKT is that researchers and knowledge users work together to solve problems by collaboratively generating knowledge. The typical IKT approach has been implemented at project, regional and national levels. Two other models of researcher/knowledge user interactions—consensus-based approaches and networks—have captured our attention under the IKT umbrella as traditionally conceptualised. Attention is also needed to identify or create methods to meaningfully engage knowledge users in the various stages of the knowledge-generation process.

Researchers and knowledge users jointly problem solve in consensus-based approaches, such as think tanks, task forces, priority setting exercises, consensus conferences, Delphi panels and the like. In each of these examples, researchers and knowledge users come together to discuss, reach a new understanding and then apply their co-developed knowledge. The outcome might take the form of a report, recommendations, practice guidelines or priorities. The nuance here is that a specific research question may not necessarily be the guiding force in these approaches. Instead, there may be a ‘pull’ related to a policy issue requiring a policy response or an uncertain practice issue requiring clarification. For example, a recent approach being used specifically by knowledge translation scholars is the deliberative dialogue,13–15 a process that brings together stakeholders (including researchers) to discuss summarised research findings and contextual evidence to influence policy. The technique is meant to highlight contested areas such that participants can come to a common understanding of the issue. Here is another example: in the area of systematic evidence reviews and practice guidelines, the Grading of Recommendations Assessment, Development and Evaluation (GRADE) methodology has emerged as a way to integrate quality of evidence with values (eg, identification and rating of patient-important outcomes, as opposed to only study-important or study-available outcomes) and context, including feasibility of intervention adoption in different settings, applicability to different groups of patients, system factors, etc. GRADE is a consensus-based process that can include stakeholders with varying views of the content and context to arrive at recommendations that are appropriate and actionable.16 Adaptations of GRADE have emerged to address a variety of topics, and GRADE is now part of the WHO's Guideline Development process, which recommends the inclusion of a variety of stakeholders on Guideline Development Groups, and on its external review panels.17 (That said, application of GRADE across the WHO's different Guideline Development Groups remains inconsistent.18) By considering these types of consensus-based approaches as falling within the umbrella of IKT, the boundaries of IKT can expand while still holding on to its underlying principle of valuing knowledge from researchers combined with knowledge from other knowledge users to address societal or practice issues.

Another common example of multiple researchers and knowledge users working together is through an interorganisational network structure, which brings together different types of people with varied expertise and knowledge. Networks are gaining momentum as governments and funding agencies provide incentives for their formation; evaluations of their effectiveness are just emerging.19 These collaborative networks focus on a large, often complex, issue, and take a multipronged approach to addressing a health or social problem.19 ,20 Network activities include researchers and knowledge users working together on such things as generating reviews of the evidence, disseminating research, developing practice guidelines to support implementation of evidence, etc. For example, The Preventing Violence Across the Lifespan (PreVAiL) Research Network, established in 2009, is an international research collaboration of over 75 researchers, trainees and knowledge users from Canada, the USA, the UK, Asia, Europe and Australia. PreVAiL members have participated in team meetings, research priority setting processes,21 research projects and mentoring of trainees. PreVAiL is an example of a non-traditional IKT model involving researchers and partners (ie, organisations comprising practitioners, policymakers and/or other types of knowledge users) that is demonstrating some promising findings in meeting the goal of research-to-action.22

Joint consensus-based approaches and networks are just two examples of many that expand the boundaries of how we think about IKT. In addition, innovative engagement strategies are required to support knowledge user input. If both researchers and knowledge users are responsible for addressing research problems, there need to be ways to initiate and sustain both parties' interest. In particular, researchers are encouraged to give thought to how knowledge users can participate in data collection or data interpretation while remaining within the scope of university ethics board requirements.

IKT tools for engagement

Some tools, derived from systems science, can be applied to IKT to enable participation by knowledge users. For example, the technique of concept mapping facilitates active involvement by knowledge users, through repeated sorting, rank ordering and visualisation techniques on their part, and cluster analysis and multidimensional scaling on the part of the software, to arrive at a systematic conceptualisation of actionable knowledge.23 ,24 Knowledge users provide input (data) based on their expert knowledge and also have the opportunity to interpret the concepts emerging from the analysis alongside the researchers. Social network analysis (SNA) is used to analyse the connections between different organisations/people in a social network; the approach can be used to understand interactions related to social support relationships, transmission networks or information sharing relationships, for example.25 ,26 SNA provides a way to identify the structure of a network, providing insight into the influence of the network structure on individuals within a group. What is important for IKT is that SNA findings can be transformed into sociograms, or easy-to-understand visual depictions of relationships that have been used for important discussions about research findings with knowledge users.27 ,28

Narrative formats can also be used to engage IKT knowledge users in the research process. Narratives tell a series of events or ‘facts’ as a holistic story, and may uncover issues not otherwise apparent, while also facilitating understanding.29 They allow the blending of facts with contextual knowledge, including local knowledge about “the setting in which practice takes place” (p. 94),30 as well as knowledge about community systems and structures that influence health and well-being.31 Understanding context as a form of knowledge requires what Contandriopoulos et al31 call a “detailed analysis of the context” (p. 468), and we have found, at least in some of our work, that narrative formats provide an excellent bridge between data-derived facts and local contextual knowledge, and facilitate common understanding. For example, as part of an IKT-based intervention study co-led by one of us (CNW), a case study design was used to implement, tailor and evaluate an intervention to increase the organisational capacity of four primary healthcare clinics to deliver more equity-oriented care to the populations (mostly people experiencing significant marginalisation) they served.32 As part of the case study design, the research/knowledge user team developed ‘clinic and community context profiles’ to describe each site's mandate, clients, history, etc. As the research evolved, we shared and discussed these profiles with the sites, and began to use them—in different versions—as vehicles to document specific activities, such as intervention components and progress, including, for example, key results from the intervention's baseline data collection phase. These data were contextualised by the team to highlight where each clinic's patients stood relative to local, regional and provincial rates for specific health indicators (eg, chronic pain, post-traumatic stress disorder symptoms). The clinical narratives came to be used by the sites in a variety of ways, from orienting and training new staff, to using the patient data in funding requests.32 ,33 While more emergent than the more structured approaches described above, in the context of this specific IKT-designed intervention study, the clinic narrative became an effective ‘knowledge intervention’ in its own right (eg, see

Knowledge users can also think about structured ways for researchers to engage with their organisations or other sites of knowledge application to activate the potential for research impact. In our experience, formal and informal presentations, meetings and other interactive opportunities for researchers to engage with the teams, departments, units, etc, of their knowledge users provide excellent spaces for understanding ‘research-in-context’.22 Knowledge users are also encouraged to self-organise as a knowledge user team rather than identifying one liaison person who connects with researchers. In this way, staff turnover is less problematic, and projects are not jeopardised on account of corporate memory loss or a need to rebuild relationships with a new staff member.

Considerations and cautions

Thus far, we have described the traditional concept of IKT accompanied by examples of other collaborative work models that may in fact fall within IKT's scope. Recent reviews of IKT studies demonstrate that there is still much to learn about what makes IKT work and what factors impede IKT progress. There is also room for the development of innovative tools that support a process whereby ‘society can speak to science and science can speak to society’ through IKT involvement. We suggest that the next major frontier in IKT evaluation is determining appropriate outcomes and impacts. In particular, we need to assess the meaning, value, effectiveness and efficiency of IKT teams from multiple perspectives, including those of our knowledge users. This includes attention to how individual IKT team members transform their understanding of a problem and solutions to a common, collective world view—Does our prediction of this transformation hold, and does it matter? Work is needed to identify the important impacts resulting from IKT processes, with the acknowledgement that impacts might be diffuse and not readily measurable. As Greenhalgh and Fahy10 argue, such impacts are “less linear, more complex and multi-stakeholder, and thus harder to demonstrate” (p. 8). The authors note that using a complex systems theoretical lens, for example, might lead to definitions of impact encompassing interactions between organisations or sectors. They also recommend the explicit consideration of power differentials and conflicts of interests, and their influence on impact, in assessing co-production models of research like IKT.

In our previous paper, we noted some risks associated with IKT, such as traditional performance evaluations that do not recognise the efforts of those working with IKT approaches. Other risks include partnership imbalances with research priorities driven by the researcher or knowledge user rather than jointly determined, leading to ineffective health practices or policies, or indifference to research rigour. There are also a number of ethical and methodological issues that researchers might find useful to consider when entering into IKT approaches to research. First, the knowledge produced by research, especially individual studies, does not always fit unproblematically with knowledge user needs. In our experience, negative or null results can prove very disappointing to those expecting research to demonstrate support for a specific intervention or programme. This can cause ruptures in the immediate IKT collaborating team, and also in broader networks of stakeholders, where results that cause dissonance can be misrepresented, partially appropriated, ignored and even vilified.35–37 Evidence and knowledge are rarely fixed and objective, and their application even less so.38

Also, while we and others have outlined the many potential benefits of IKT approaches to addressing research–practice gaps and discontinuities, one potential limitation for researchers, and for science and society more broadly, is that the entire IKT approach is premised on the fact that knowledge users come to the research process with specific needs for evidence that will address their already-existing problems. In effect, this can serve as a barrier to true innovation in that major breakthroughs that require paradigm shifts are unlikely to be pursued. The more that funding agencies require partnered, IKT-modelled research projects, the less room there is for curiosity-driven, potentially ‘game-changing’, research. This might be especially true in industry-sponsored or industry-partnered initiatives (as in public–private partnerships), where success is defined primarily in immediate economic terms.


In this paper, we raise important areas of investigation for IKT scholars, that is, development of joint data collection and interpretation tools, as well as the identification of conceptually relevant outcomes. Some of this work will be carried out by the Integrated Knowledge Translation Research Network, a 7-year federally funded research programme in which the first author and 40 other researchers/knowledge users are involved. When deciding how to define ‘success’ in IKT approaches, and selecting metrics to measure these, we must guard against setting ourselves up for failure—what can an IKT modelled project hope to achieve in the immediate and longer term? How can these outputs and outcomes be assessed and linked? Given the many and varied contextual factors that influence the entire knowledge-to-action process, it is unclear what the IKT aspect of a project can meaningfully claim as a unique or significant contribution to outcomes. Similarly, further discussion is required on the topic of unintended consequences, or even ‘IKT failures’. Finally, IKT researchers and knowledge users need to consider what is given up when deciding to use an IKT approach, and how to assess the balance of potential benefits and costs (including direct and opportunity costs) of such models? Clearly, much more work is still needed to clarify and enrich the concept of IKT.


Selma Tobah provided manuscript formatting support.


View Abstract


  • Contributors AK and CNW jointly conceptualised the work. AK led the development of the manuscript. All authors read and approved the final manuscript.

  • Competing interests None.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.