The effectiveness of knowledge-sharing techniques and approaches in research funded by the National Institute for Health and Care Research (NIHR): a systematic review

Background The National Institute of Health and Care Research (NIHR), funds, enables and delivers world-leading health and social care research to improve people’s health and wellbeing. To achieve this aim, effective knowledge sharing (two-way knowledge sharing between researchers and stakeholders to create new knowledge and enable change in policy and practice) is needed. To date, it is not known which knowledge sharing techniques and approaches are used or how effective these are in creating new knowledge that can lead to changes in policy and practice in NIHR funded studies. Methods In this restricted systematic review, electronic databases [MEDLINE, The Health Management Information Consortium (including the Department of Health’s Library and Information Services and King’s Fund Information and Library Services)] were searched for published NIHR funded studies that described knowledge sharing between researchers and other stakeholders. One researcher performed title and abstract, full paper screening and quality assessment (Critical Appraisal Skills Programme qualitative checklist) with a 20% sample independently screened by a second reviewer. A narrative synthesis was adopted. Results In total 9897 records were identified. After screening, 17 studies were included. Five explicit forms of knowledge sharing studies were identified: embedded models, knowledge brokering, stakeholder engagement and involvement of non-researchers in the research or service design process and organisational collaborative partnerships between universities and healthcare organisations. Collectively, the techniques and approaches included five types of stakeholders and worked with them at all stages of the research cycle, except the stage of formation of the research design and preparation of funding application. Seven studies (using four of the approaches) gave examples of new knowledge creation, but only one study (using an embedded model approach) gave an example of a resulting change in practice. The use of a theory, model or framework to explain the knowledge sharing process was identified in six studies. Conclusions Five knowledge sharing techniques and approaches were reported in the included NIHR funded studies, and seven studies identified the creation of new knowledge. However, there was little investigation of the effectiveness of these approaches in influencing change in practice or policy. Supplementary Information The online version contains supplementary material available at 10.1186/s12961-024-01127-5.


Background
Academic research has little influence on the commissioning, design and delivery of health care services [1][2][3].Stakeholders, including patients, are currently not consulted sufficiently for research to be genuinely informed by their experiences [4,5].This is of concern to research funders globally, who have a remit to fund health and social care research that improves people's health and wellbeing [6].Knowledge mobilisation is a generic term that refers to making knowledge ready for action and includes activities ranging from dissemination to coproduction [7].Other similar terms are often used such as knowledge translation, knowledge exchange and integrated knowledge translation (IKT).For the purposes of this review, the key element of knowledge sharing was focused on within the field of knowledge mobilisation to explore knowledge mobilisation as an intervention and an active process, within research studies.Exploration of the lack of integration between researchers and stakeholders within the fields of knowledge mobilisation and implementation has highlighted that knowledge sharing needs to be a two-way process and not, as previously accepted, a linear one [8][9][10][11].This shift in understanding has been driven through a recognition of the complexity and messiness inherent in bringing together different communities to develop a common or shared understanding [3,12].Consequently, activities to improve knowledge sharing and implementation have shifted away from targeting research findings towards patients, practitioners and policy makers and been replaced with techniques to encourage two-way knowledge sharing and co-production [9,[13][14][15].A variety of theories, models and frameworks have been used to support this two-way process, with varying degrees of success [16,17].
Knowledge mobilisation is defined by the NIHR as 'sharing knowledge between different communities to create new knowledge to catalyse change' [18].There is consensus that if knowledge is shared between two or more communities, it can result in the creation of new knowledge, which has a greater likelihood of leading to change within practice or research [7,[19][20][21].Change that can be linked back to original research findings or outcomes is often referred to as research impact [22][23][24].Techniques and approaches that have been developed to follow this mechanism of knowledge sharing include, models of embedded researchers or practitioners, use of knowledge brokers, stakeholder engagement, organisational collaborative partnerships and the involvement of stakeholders in the research or service design process itself.For example, embedded models can facilitate the knowledge sharing process by a researcher or health care practitioner leaving their home organisation to work in a host organisation, thereby increasing the opportunities for sharing knowledge between the two organisations.The underlying premise is that it is through people and their interactions that knowledge is shared and by increasing the proximity of individuals this can facilitate interactional opportunity [10,25,26].They may be hosted by one organisation, but their function is to work between the organisations to facilitate knowledge sharing [27][28][29].Stakeholder engagement, when conducted for two-way knowledge sharing, involves inviting stakeholders to share knowledge at specific meetings, workshops and events [30].Involving stakeholders in the research or service design process as equal decision makers, advisers and informed representatives of their community, can also follow two-way knowledge sharing [21,31,32].An additional mechanism is knowledge sharing at an organisational level, where collaborative partnerships are formed [33].
In the United Kingdom, the National Institute of Health and Care Research (NIHR) awards around £1 billion in research funding per year and, along with other funders, has a strong remit to reduce the research to practice and policy gap [34].Yet, to date, there has been limited research that systematically explores and identifies the knowledge sharing techniques and approaches in the NIHR portfolio of research studies.One review examined the mechanisms and pathways to impact of NIHR funded public health research (Boulding, Kamenetzky et al. 2020).It explored the mechanisms and pathways reported on Research fish (a database for researchers to document impact related activities) and triangulated this with qualitative data exploring the researchers' perspectives of the impact of their research.The authors concluded that the standardised measures were not capturing impact in localised settings or longer-term impact [23].A second study explored the public health researchers' perspectives on impact reporting and highlighted a need for funders to identify their expectations of the impact resulting from the research they fund and to increase their support for knowledge mobilisation activities [24].These studies highlighted the need for researchers to have a clearer understanding of the knowledge mobilisation techniques and approaches to inform pathways to impact and focused on NIHR health funding streams [23,24].To our knowledge, there has been no systematic review that describes the knowledge sharing techniques and approaches that have been applied in NIHR funded research nor synthesises their effectiveness.
This review aimed to answer the following questions: (1) Which knowledge sharing techniques and approaches have been included in NIHR funded health research?(2) How effective are these knowledge sharing techniques and approaches in creating new knowledge that can lead to changes in practice and research?

Methods
The protocol for this systematic review was registered on the International Prospective Register of Systematic Reviews (PROSPERO, CRD42020171293; reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses [35]).A restricted systematic methodology was chosen to balance methodological rigour with the resources available [36].

Search strategy
Electronic databases MEDLINE via OVID and The Health Management Information Consortium, which is a compilation of data from two sources, the Department of Health's Library and Information Services and King's Fund Information and Library Services, were searched from inception to 24.4.20 for published studies, which was then updated and rerun on the 1.7.22.The search strategy was based on the terms for the intervention (knowledge sharing techniques and mechanisms, including terms for knowledge transfer, exchange and translation) and population (researchers with patients, clinicians or health services managers) (Additional file 1: Search Strategy).Additional references were identified from reference lists of included full papers.

Eligibility criteria
This systematic review included studies that described knowledge sharing between researchers with patients, members of the public, clinicians, health service managers (i.e.commissioners, policy makers and hospital managers) or voluntary agencies, that were funded by the NIHR (Table 1).Knowledge sharing was defined as 'any interactional activity through any medium (including in person, email, telephone, etc.) that involves knowledge sharing about healthcare' .For the purposes of this review, knowledge sharing techniques and mechanisms were considered as an intervention, i.e. 'the act or an instance of intervening' [37], where an explicit knowledge-sharing approach had been adopted in contrast to the established process of knowledge remaining within one community.The setting was defined as any healthcare setting, e.g.primary, secondary, tertiary health care services and public health.The outcome was defined as the use of evidence in policy and practice or the involvement of stakeholders in the research process.Where relevant, studies were included irrespective of comparator group.All study designs were included, except protocols and reviews of literature.Only studies published in the English language were included.Studies were excluded if they did not describe knowledge sharing between researchers and a stakeholder group, e.g.describing knowledge sharing between two other stakeholder groups (e.g.clinicians with health service managers, clinicians with patients and patients with health service managers).

Study selection
Records were exported and deduplicated in Endnote and then imported to Covidence for screening [38,39].The title and abstract screening was conducted by one reviewer (H.B.), with a 20% sample independently screened by one of two reviewers (C.T. and R.D.).Any discrepancies were resolved by discussion.A third reviewer (A.H.) arbitrated if needed.Full text screening

Data extraction
Data from included studies were abstracted by one reviewer (T.S.) into a data extraction form, which was piloted a priori on 10% of the included studies (S.P.) and checked for accuracy by a second reviewer (H.B.).Extraction included: study design, author name, author, year, aims, population, intervention/approach and a detailed intervention description.In some instances, studies contained a knowledge sharing element, which was not the primary focus or outcome of the study.In these cases, the detailed description of this element of the study was extracted as the technique or approach.A modified template of the TiDieR checklist was used [40].The data were extracted on the design, presence of an evaluation, use of theory or goal, procedures, materials used, context influencing factors, tailoring modifications and assessment of outcome and applicability.

Quality appraisal
Quality appraisal was conducted independently by T.S. with a 20% sample of included studies, which were reviewed by H.B., followed by discussion for any discrepancies.The Critical Appraisal Skills Programme (CASP) qualitative checklist, Critical Appraisal Skills Programme (2018) [41] was used where appropriate.The CASP qualitative checklist includes two screening question (yes/no) and an additional eight questions (yes/ no/can not tell) if the response to both screening questions were 'yes' .As outlined by Long and French, the quality of studies was assessed with a focus on the rigour of the data analysis, with consideration of the trustworthiness of the results given [41].Using this focus with the overall score from the checklist, the studies were categorised to be of high, moderate or of lower quality.

Data synthesis
A narrative synthesis method was adopted, as it includes a formal analytical process of synthesis to generate new insights [42].This narrative synthesis focussed on four key elements: (1) identification of a theory of change.In this review, knowledge sharing as a mechanism to facilitate change was used to explain the anticipated process.
(2) Development of a preliminary synthesis of the findings of included studies.A preliminary synthesis was conducted to organise the results of the included studies and identify any factors that influenced the results reported.This was conducted by developing initial descriptions of the results of the included studies, which were then organised to describe patterns, so that the factors impacting on the mechanisms of the intervention could be identified.(3) Exploring relationships in the data.The studies were explored for relationships within and between studies, which involved a process of concept mapping supported by qualitative case descriptions.In particular, the studies were examined for instances where similar mechanisms may be at work even though the overall approach may be described differently.This process was initiated by H.B. in categorising the data under overarching themes based on the mechanism of knowledge sharing, which were refined further through discussion and reflection with L.B. and T.S. into subheadings.
(4) Assessing the robustness of the synthesis.An assessment of the robustness of the synthesis was made and only studies that reached a minimum standard of methodological quality assessed by T.S. were included in the final synthesis [43].

Study characteristics
The characteristics of the included studies are shown in Table 2.These were the author, year, aims, population, knowledge sharing technique or approach, mechanism of knowledge sharing and outcome (new knowledge or change in practice or research).

Quality appraisal
Five of the included studies were descriptive studies and could not be included in the quality appraisal process [20,49,51,53,57].Of the remaining 10 studies, two were rated of moderate quality [47,48] and eight were rated as high [44-46, 50, 52, 54-56].Two studies could not be rated as they provided insufficient detail on the knowledge mobilisation intervention, so these were excluded from the final synthesis (Table 3).

Types of knowledge sharing techniques and approaches
Five explicit forms of knowledge sharing studies were described in the included studies (Table 2).Three studies applied embedded models of researchers or practitioners [20,44,54], and two studies used knowledge brokering.[46,47].Stakeholder engagement approaches that applied two-way knowledge sharing were used in five studies.These were either priority setting consensus building workshops [51,55,57] or facilitated knowledge-sharing events [49,52].Three studies described approaches where non-researchers were involved in the research or service design process itself.One study did this with patients and members of the public in research projects and another with professionals [53,56].The approach of involving patient and public members was also used in another study to assist with service design [45].Two studies examined organisational collaborative partnerships between universities and healthcare organisations [48,50].

Timing within research cycle
Six studies applied a knowledge-sharing approach to topic identification [44,46,47,49,50,55], and one study extended topic identification to also defining the research question [57].Five studies used a knowledge-sharing approach for the conduct of the research [20,48,53,54,56].One study used knowledge sharing to facilitate the adoption of findings [52], and two studies used knowledge sharing for the production of service design [45,51].There were no studies that used a knowledge-sharing approach or technique for designing the research or preparing the funding application.

Sources of NIHR funding
Eight of the studies were funded or supported by a Collaboration for Leadership in Applied Health Research (CLAHRC) [44,46,48,50,51,[54][55][56].One study was funded by a Knowledge Mobilisation Research Fellowship [45], and one study reported support from both a Knowledge Mobilisation Research Fellowship and a CLAHRC [20].Two studies were from the Health Services and Delivery Research funding stream [49,53], one study was from multiple sources, including NIHR funding [47], one was funded by the Public Health Research Programme [52] and one was funded by Programme Grants for Applied Research Funding [57].

Use of theory
Of the 15 studies, 6 studies drew upon or referred to a theory, theoretical basis or used a framework [20, 46-48, 50, 56], (Table 4).The theory most frequently drawn upon was that of Communities of Practice [60,61], which     was referred to by three of the studies to explain the process of knowledge sharing [20,50,56].Two studies drew upon other theories to explain knowledge sharing as part of a co-production process.One referred to Ritual Theory [62] and the concept of Interaction Ritual Chain [56,63], and the other used three theoretical lenses, the coproductionist idiom [64], interactionist currents within organisation studies [65,66] and communication, argumentation and critique from a pragmatic perspective [67,68], In Ref. [48].Another study drew on the sociological theory of dramaturgical perspective [47,69], and one study used the frameworks of why, whose, what and how [70] and PAHRIS [71] to explain their approach [46].Only one study explicitly referred to a theory of change and outlined a potential process [50].Nine studies did not use any theory or frameworks to explain or predict the knowledge sharing process leading to change [44,45,49,[51][52][53][54][55]57].

Knowledge sharing as a mechanism to facilitate change
The theory of change identified from a preliminary synthesis of the included studies followed the process outlined within the literature, which is shown in Fig. 2.
All studies confirmed the causal direction of the knowledge sharing mechanism as shown by the arrows in Fig. 2 and were found to be following the process of knowledge sharing across communities with an intention of creating new knowledge (Table 2).Seven studies reported that new knowledge had been created through knowledge sharing [45,50,51,[53][54][55]57].However, only three studies attempted to outline the anticipated change from the knowledge-sharing approach [45,50,53], and only one study provided any evidence of change [54] (Table 2).

Evaluation of knowledge sharing technique or approach
Ten studies conducted an evaluation of the knowledge sharing technique or approach to understand its process or effectiveness (perceived or intended) [44-48, 50, 52, 54-56], (Table 4).The other five studies gave detailed descriptive accounts of the knowledge sharing process [20,49,51,53,57].There was no relationship between the knowledge-sharing approaches used and whether an evaluation was conducted.Three studies using stakeholder engagement approaches gave a process description [49,51,57], one involvement study [53] and one study using an embedded model [20].Of those studies that conducted an evaluation a range of methodologies were used, which were predominantly qualitative.Six studies used semi-structured interviews [44,50,52,[54][55][56], three studies used mainly observational methods [48,52,56], two studies used document analysis [55], two studies used reflective diaries [44,46] and two studies analysed field notes and emails or meeting recordings [45,47].Other methods used were focus groups, surveys and postal questionnaires [45,48].Five of the studies that conducted an evaluation of the knowledge sharing technique or approach drew upon a theory or framework to understand or explain the process [46-48, 50, 56] (Table 4).

Evidence of effectiveness
Of the seven studies that reported the creation of new knowledge [45,50,51,[53][54][55]57], four also evaluated the process and also attempted to outline the anticipated change from the knowledge-sharing approach [45,50,54,55].One of these studies used the knowledge-sharing approach of involvement of stakeholders in service design, one explored an organisational collaborative partnership, another used an embedded model and the other a stakeholder engagement approach [45,50,54,55].The only study that reported a change in practice or research did not outline the process of change and did not explain the process using a theory or framework [54].However, this study of an embedded model was the only report of a change in practice as a result of a knowledge sharing technique or approach (Table 4).

Discussion
This review summarises the knowledge sharing techniques and approaches used in NIHR studies between 2006 and 2022.Five knowledge sharing techniques and approaches have been included in NIHR funded health research: embedded models, knowledge brokers, stakeholder engagement, involved research or service design and organisational collaborative partnerships.In applying a mechanism of knowledge sharing, three studies outlined anticipated change from the process of knowledge sharing using the approach of stakeholder involvement [45,53] and organisational collaborative partnerships [50], and only one study provided evidence of change, which used an embedded model [54].
We found that in some studies knowledge sharing techniques and approaches were used but not identified using New knowledge that could not have been developed without the involvement of more than one community.

Potential to catalyse change within research or practice
Knowledge sharing across more than one community.
Fig. 2 Theory of change model developed to inform initial synthesis established terminology and in other studies terminology was used interchangeably, with a lack of consensus on the definition of terms.This may well reflect the developments overtime in how knowledge is mobilised in a nonlinear fashion, as this review included papers from 2008 and tracks the gradual establishment of agreed terminology.However, a current lack of clarity of terms has been identified in the literature around co-design, co-production and co-creation, where terms are used interchangeably and clarity around the aims of the approaches are unclear [72].This seems also to be the case in what we have referred to as the embedded models, which included researchers in residence and secondment opportunities.It was unclear in synthesising the studies what the different roles were that these terms applied to, as terminology was used differently across the models for example using the term knowledge broker to refer to an embedded researcher working within clinical practice [44].Knowledge sharing techniques and approaches were often used without reference to underlying theory or an explanation of the anticipated change process.Although an acknowledgment of the clarity provided by a clear theoretical basis to understand the process of knowledge mobilisation has been accepted, this has been relatively recent [73,74].Recent studies have highlighted and categorised a large number of theories, models and frameworks available but acknowledged a limited evidence base on their use [75,76].In this review, only six studies drew on a theoretical base to explain or predict causality, and only four studies used this for evaluating the knowledge sharing technique or approach.A recent systematic scoping review of knowledge transfer and exchange models also noted a lack of evaluation of the processes and outcomes by those engaged in knowledge mobilisation activities [77].Evaluation models do exist in the field that construct a framework for assessing impact or change at multiple levels, which also take account of the inherent complexity and uncertainties in assessing change [7].To encourage greater use of knowledge mobilisation techniques and approaches amongst non-specialists, more explanation of these is needed to facilitate replication with confidence.Studies describing a knowledge sharing technique or approach without reference to an output, outcome or change mechanism, risk losing the interest of the wider research community, as the benefits of this approach are unclear.
This review included studies where knowledge sharing techniques or approaches could be identified but may not necessarily been acknowledged by the authors.Where knowledge-sharing approaches were not acknowledged, the knowledge sharing component was often not reported in detail.For example, in Batchelor 2013, the knowledge-sharing element of the James Lind Alliance Priority Setting Partnership was given little attention in the reporting and was difficult to untangle from the information gathering element of the study [57].As an older study this may reflect less interest at the time in the process of knowledge sharing with stakeholders, although there were clear attempts to extend the remit of the James Lind Alliance to include researchers in the workshops and to involve stakeholders in designing the research questions.Unfortunately, the lack of detail on the procedure reduces the opportunity for replication or wider evaluation when a project is deemed to be successful, reducing the opportunity for future learning.In work involving public contributors, researchers often gave a more detailed account of process and procedures, which may indicate greater maturity in the field for working with this stakeholder group.This may also give an indication as to why so few studies reported on their knowledge sharing activities and intended impact.As the request from funders for the demonstration of research impact is a relatively new requirement, previous work in this area may not have been seen as important or as a core component of a research study.Likewise, prior to the agreement from funders to fund and support impact related activities such as knowledge mobilisation, achieving impact in services or society may have not been seen as within the remit of the research community to deliver.
Promising techniques and approaches that were evaluated, often focused more on acceptability of the approach rather than whether new knowledge was created.This may have been due to an interest in how to maintain ongoing work with stakeholders, or possibly a lack of confidence in the technique or mechanism leading to new knowledge or in the sensitivity of the evaluation to identify it.Although knowledge sharing can be seen as a simple concept, achieving an authentic approach is known to be a complex process [7,78].It is not to suggest that complexity does not exist, only that current reporting may render the purpose of knowledge sharing techniques and approaches invisible to those outside the specialist field.While the importance of identifying and reporting on impact remains a central issue to funders, identifying techniques and approaches that can lead to changes in practice and research will be of value.Currently the NIHR as a funder, requests engagement and impact plans in applications for funding and advocates the use of knowledge mobilisation strategies from the outset of the study to achieve this [18,79].Monitoring of the impact from NIHR funded research is then conducted for 5 years after study completion via an online system (Researchfish) [80].

Strengths and limitations of the review
This systematic review restricted the number of database searches to two and did not explore grey literature, which may have resulted in not identifying all relevant studies.The included studies were also restricted to the English language.However, given that this review is focused on the literature produced by the major UK funder with a requirement for publication in mainstream open access journals, this is less of a concern.A restricted systematic review methodology was used to balance rigour with the resource available [36].This requires only a proportion of the screening, full-text review and data extraction to be conducted by two reviewers.Given the difficulties with the terminology, unclear methodologies and complex study designs, studies may not have been identified through the initial searches.As outlined earlier, studies often did not report knowledge mobilisation or knowledge sharing activities in a thorough way and this led to difficulties with data extraction and may have led to an underestimation of use of knowledge-sharing approaches.This review specifically focused on the relationship between knowledge sharing as a key element of knowledge mobilisation activity, leading to the creation of new knowledge with the potential to lead to changes in practice or research (impact).Studies that mobilised knowledge for other outcomes were excluded, which may be a weakness in understanding knowledge mobilisation processes more generally.A key strength of this review was the attempt to apply a robust review framework to an often-confusing field of terms and mixed approaches.An established framework was applied to synthesise the current knowledge in this field with the intention to collate the learning to date and to guide those who are not specialists in knowledge mobilisation towards the techniques and approaches which might be useful for future research.

Key learning
There is a need for clear reporting in the field of knowledge mobilisation that recognises the goals of these techniques and approaches.Theories and models exist that support exploratory work and complex systems, which could be used more widely to explain the knowledge sharing mechanism of knowledge mobilisation approaches.Evaluations of these techniques and approaches could be better linked to the underlying goals or outcomes of change and impact via established theories and explanatory models.This would enable researchers not specialist in the field of knowledge mobilisation to better understand the field and have confidence in introducing these techniques and approaches into their work.Clearer reporting on knowledge sharing processes and outcomes can support the research community and funders alike in identifying where knowledge mobilisation can assist in closing the research to practice gap.

Conclusions
There is little evidence of the effectiveness of knowledge sharing techniques and approaches used in NIHR research studies in influencing change in practice or ongoing research.This does not mean these techniques and approaches are not effective in instigating change or impacting on practice, rather that clear evidence for this has not yet been produced.Although a complex and often messy field, there are theories, models and frameworks that can be used to shed more light on techniques and approaches that currently show promise but lack evidence for their effectiveness.

Table 1
Inclusion and exclusion criteria Outcome of interestPrimary -relevant techniques or approaches to inform the practice of knowledge sharing Secondary -have been deemed successful or notDesignTo explore how a technique or approach is working.Either detailed description or an additional methodology that explores the processes of the technique or approach was conducted by one reviewer (H.B.) with a 20% sample independently screened by one of two reviewers (T.S. and L.B.); any discrepancies were resolved by discussion.A third reviewer (S.R.) arbitrated if needed.

Table 2
Characteristics of included studies

Table 3
Quality appraisal of studies

Table 4
TIDieR intervention checklist