Introduction

Health services are under continuing pressure to improve the quality of care they provide and the outcomes of such care for patients. To address this, health services employ a wide range of strategies such as the implementation of new treatments, devices and models of care, the recruitment of skilled staff, professional development of existing staff, investment in quality improvement staff and initiatives, development of IT and other care delivery support systems, tools and resources, and establishment of care delivery and performance monitoring systems [1–5].

The conduct of research in health services, or research engagement by health service providers, has also been suggested to have additional benefits to the quality of patient care and outcomes through indirectly modifying the culture and practice of care delivery [6, 7]. Research engagement has been variably defined with one review describing clinician research engagement as a ‘deliberate set of intellectual and practical activities undertaken by health care staff and organisations to conduct research’ [8]. Systematic reviews have reported improvements in health outcomes and provision of care where clinicians and organisations are research-engaged [6, 7]. Further, several longitudinal studies have reported that provider and/or health care organisation participation in clinical trials can result in lower mortality rates and greater adherence to clinical guidelines [9]. This greater focus on ‘knowledge production/generation’ by health service end users, and engagement of such end users in identifying research priorities has been suggested to promote better quality health care and patient outcomes [10]. In response, a shift in health and medical research and health service delivery policies toward increasing the engagement of health care services in the conduct of research is occurring internationally [11–13].

One example of this is  the National Institute for Health Research (NIHR) in England where 10 million pounds was invested to establish nine collaboration units (Collaborations for Leadership in Applied Health Research and Care (CLAHRCs)) in 2008 aimed at engaging NHS staff to participate in research to improve patient outcomes [11, 12]. A number of funding structures have been put in place to support the integrated Knowledge Translation (iKT) research model in Canadian hospitals to increase researcher’s collaboration with ‘knowledge users’ (i.e. health care providers/users) to increase the generation of evidence that is more readily able to be implemented into practice [14]. In Australia, over $10 million was invested in building Advanced Health Research and Translation Centres (AHRTC) and Centres for Innovation in Regional Health (CIRHs) to facilitate collaboration between researchers, health care and industry stakeholders to undertake applied research directly relevant to improve patient care [13].

Identifying the research question

 Although healthcare professionals, policy makers and the public at large recognise that research engagement by health care organisations is worthwhile, little is known about the evidence base surrounding research engagement and the type of strategies that have been used to facilitate increased research engagement. Previous reviews examining this have examined research engagement together with how to increase clinicians’ application of research in practice (i.e. implementation science). A review by Boaz et al. examined potential mechanisms between clinician engagement with research and improved health service and patient outcomes [6]. The review found that supporting clinicians to undertake research was a distinct behaviour from increasing clinicians’ implementation and multifaceted strategies are required to support research engagement, findings consistent with other reviews describing research co-production. For example, a review by Gagliardi et al. provides a summary of how integrated knowledge translation approaches have been applied to target decision makers’ engagement with research and described a range of individual and structural barriers and enablers that need to be targeted to support research engagement by health care organisations [15]. The previous reviews examining clinician research engagement specifically however have focused on describing the impact of capacity building strategies that target individual factors only [16, 17]. Our review seeks to extend on this literature by providing an overview of the different strategies that can be used to promote research engagement by health care providers [16]. In areas such as this where terminology, types of strategies and outcomes are unclear, an examination of the volume and characteristics of available research is needed to provide an overview of the evidence available to inform practice and policy making, and to assist in identifying gaps and areas for future research investment to increase health provider/ organisation research engagement [18]. Scoping reviews are recommended to provide an opportunity to clarify key concepts; describe gaps in the research; and characterise the types and sources of evidence to inform practice, policymaking and future research [18].

Objectives

As such, to identify gaps in the area and opportunities for further research, this study aimed to: i) describe the percentage, type and study design of publications that examined research engagement by health care providers and organisations and ii)provide a list of the types of strategies used to promote research engagement by health care professionals and organisations reported in the literature.

Methods

Study design

A scoping review using systematic search processes was undertaken to locate peer-reviewed publications and grey literature related to ‘research engagement by health care providers and organisations’. The manuscript was reported consistent with the PRISMA-ScR [19]. No protocol was registered; however a copy can be provided from the corresponding author upon request. The definition of research engagement applied in this study was informed by a comprehensive commissioned review undertaken by Hanney et al. (2013) which distinguishes between ‘engagement with research’ and ‘engagement in research’ with the former including a less substantial involvement at an organisational level and relating more to receiving and transmitting research findings, more conventionally known as knowledge translation or implementation (evidence-based practice) in the literature [8].

Similar to the review by Hanney et al., this study focused on ‘engagement in research’, defined as a ‘deliberate set of intellectual and practical activities undertaken by health care staff and organisations to conduct research’ [8]. This definition requires health care providers, services and organisations to have a more involved role in the entire research process including in its design, commissioning, conduct and dissemination [8].

This review applied methods consistent with the Cochrane Handbook for conduct of scoping reviews, and was based on the updated framework outlined by Peters et al. [20, 21]. This involved; i) identifying the research question; ii) identifying relevant studies; iii) study selection; iv) charting the data; and v) collating, summarizing and reporting results.

Systematic search processes and broad eligibility criteria were developed by the research team apriori, however consistent with recommendations for the conduct of scoping reviews, the search was supplemented with additional articles based on advice/recommendations from policy makers and experts in the field known to the research team, as well as additional grey literature searches using search terms identified from the initial electronic search (see Additional file 1).

Identifying relevant studies

A database search of electronic records was performed in the following electronic databases with no limit on date, MEDLINE (1950-); EMBASE (1947 -)’ PsycINFO (1950-); Academic Search Ultimate; CINAHL on the 27th December 2019.

As research engagement is not homogenously defined in the literature and consequently not always clearly indexed in electronic databases, the search strategy was developed based on extensive consultations and piloting of search terms with a university health systems librarian. The librarian, acting in line with Peer Review of Electronic Search Strategies (PRESS) 2015 Guidelines, assessed all aspects of the search terms, subject headings, search strings, limits and filters to ensure they were conceptually and functionally accurate in relation to the research question [22]. This strategy was adapted from that previously employed in a comprehensive review examining the association between research participation/engagement and health outcomes and refined to suit the purposes of this review [8]. The search strategy was kept intentionally broad, consisted of search terms for research engagement combined with search terms for health care providers/organisations. The reference list of all included studies were screened to identify further studies that could be relevant. Subsequently, a search of the grey literature using Google was undertaken using key terms identified from articles included from the electronic database search. Five separate searches were undertaken and included entering ‘research capacity building, collaborative activities/research, health research capacity strengthening and research engagement’ into the Google Scholar search engine and reviewing the first 100 hits in June 2019. Four policy makers and research experts provided additional articles that were relevant to the review aims.

Study selection

Inclusion and exclusion criteria

The main concepts of the review question which shaped the inclusion and exclusion criteria for the search can be found in Table 1.

Table 1 PCC (population, concept, context) elements of the review question
Full size table

Studies that examined research engagement were included regardless of study type. Articles not published in English were excluded from the study. Studies that examined appraising or applying the evidence solely without actual conduct of research, those that examined ‘how’ to move research into practice (i.e., implementation science) and dissemination of research findings exclusively, as well as those undertaken in non-health care settings/or with non-health care practitioners were excluded.

Screening

The search, screening and data charting processes were systematic and undertaken by two independent authors. Abstract and full text screening was undertaken independently by two authors (SY/KB/MF/AG/MK) using COVIDENCE [23]. Where there was disagreement, a third reviewer resolved any differences (LW). All studies regardless of study type were included as long as they explored research engagement (defined above).

Charting the data

Charting data from included publications was conducted according to the data classifications outlined below and consistent with previous studies [24–26].

A standardised data charting tool (see Additional file 2) was developed to chart the following additional information: author, year of publication, country where research was conducted (or first author affiliation if review), study design, study type, study setting and the detail of the intervention (verbatim from the text). A detailed instruction sheet with how to code each study was developed to ensure standardised data charting. All extraction were undertaken by two authors in Microsoft Excel, and all differences resolved via discussion (KB/SY/JW/CW).

Publication classification

All studies were categorised as either databased (i.e. data-driven, presenting data or new analysis of existing data relating to research engagement) or non-databased (i.e. not data-driven, no data or new analyses), as a means of differentiating between the different levels of evidence [27]. Databased publications were further classified according to study type, study design, translation levels and study setting.

Study type: Databased studies were further classified into quantitative or qualitative based on that reported by the authors. Where this was not explicitly described, we based this on the type of analysis used. Studies that used both quantitative and qualitative methods were classified as mixed methods. Databased studies were also categorised into measurement, descriptive or intervention research consistent with definitions from previous studies [24–28]. The following definitions were used:

  1. 1.

    Measurement: Papers developed or examined the qualities of a measurement instrument such as reliability, validity, or acceptability. Data collection methods included the use of questionnaires, interviews, physiological assessments, risk screening and observations. Papers that focused on both measurement and descriptive issues were coded as measurement research.

  2. 2.

    Descriptive: Papers exploring the frequency, patterns, correlates or predictors of research engagement strategies. These include epidemiological studies examining frequency or patterns of risk factors and correlates of research engagement.

  3. 3.

    Intervention: Papers that tested the effectiveness of an intervention to increase research engagement and/or providers. Intervention publications will be defined by the research aims rather than the study design or type of intervention. Papers that focused on both descriptive and intervention issues will be classified as intervention research.

Non-databased articles were categorised as below consistent with previous studies [27]:

  1. a

    Discussion papers or commentaries: Editorials, comments, letters, news or interviews. These articles did not present original data or describe a specific research project or intervention.

  2. b

    Case reports: Articles in which the publication description indicated that it was a case report.

  3. c

    Program description: Descriptions of methods or processes undertaken for research engagement. This category included articles that described an intervention or health initiative being applied, or that had the potential to be applied, but in which no data-based evaluation was reported.

Study design: Those with an intervention focus were further classified according to the levels of evidence (including cross-sectional, case control, cohort studies, non-randomised controlled trial, quasi randomised controlled trial, and randomised controlled trial) [29].

Study setting: The study setting was recorded as either hospital(s) or a department within hospitals, community health settings, primary care, health care setting more broadly (not specific), and networks or collaborations at a national or international level or other.

Qualitative synthesis

To produce a list of strategies used to promote research engagement (i.e. research engagement strategies), we undertook a qualitative synthesis of data from databased intervention studies, and all non-databased studies, using a Framework Approach (FA) [30, 31]. This was undertaken on the narrative description of the research engagement strategies that were extracted verbatim from the relevant studies. The FA is a type of systematic, transparent qualitative content analysis composed of several inter-related stages that allow for a structured case- or theme-based analysis of data [30, 31]. FA is conducted by a team, where analysis is aided by collaborative development of highly structured outputs (charts/displays) providing an accessible visual reference that can be interpreted by people with little qualitative research experience [31]. Such qualitative methods of synthesising review findings have been recommended to generate broad findings/key themes and to highlight opportunities to advance the evidence-base [32, 33].

As part of this, the team (JW, KB, SY, CW) employed five stages of the analysis, including; 1) familiarisation with the data; 2) identification of recurrent and important themes; 3) coding; 4) grouping codes into an analytical framework; and 5) applying an analytical framework to data [31]. The entire process was facilitated by a skilled qualitative researcher (KB). The focus of the analysis was to produce descriptive accounts to answer predefined research questions (elements of research engagement strategies that are intended to, have been used to, or have been evaluated in the context of improving research engagement by health care providers and organisations). The analytical framework(s) emerging from the qualitative data analysis was intended to be pragmatic in nature, tailored to produce information that can be used to inform public health practice. Aligned with the project specifications and aims, we applied a positivist approach to qualitative synthesis, where we comprehensively searched for empirical findings and then accumulated and described these findings [32].

Results

Collating, summarising and reporting results of included studies

From the electronic search, 13,983 references were screened by two reviewers and 53 were identified through other sources. A total of 13,996 articles were screened after removal of duplications. A total of 271 articles were included in the full text screen and 152 were included in this study (see Fig. 1 for PRISMA diagram). This total included 107 from the database search, 40 searching from the grey literature included articles and and five based on expert recommendations, resulting in a total of 152 records for inclusion (see Fig. 1).

Fig. 1
figure 1

Flowchart outlining number of studies included at each review stage

Full size image

A range of terminology was used to describe clinician and health care organisation research engagement including building research capacity, collaborative research, academic clinicians, building research infrastructure in health care, research participation, translational research, fostering clinician led research, exchanging knowledge, research networks, practice-research collaboration, integrated knowledge translation, research-based practice and research co-production.

Most studies were conducted in or had a first author located in the United States (n = 46), Australia (n = 41), the United Kingdom (including England, Ireland and Scotland, n = 36) and Canada (n = 17). The remaining were in New Zealand (n = 2), Norway (n = 1), Denmark (n = 1), the Netherlands (n = 1), Sweden (n = 1), Spain (n = 1), Switzerland (n = 1), France (n = 1), Ireland (n = 1), Japan (n = 1), and two in low to middle income countries.

Publication classifications

Of the 152 included records, 68 (44%) presented new, synthesised data (databased) and the remaining 84 (56%) were non-databased. Twenty-four studies were published before the year 2000, 46 were published between 2000 and 2009 and the remaining 82 were published after 2009 (2010–2019) (see Table 2).

  1. i.

    Study type

Table 2 Study characteristics for all included studies *two studies are missing from this table, apologies for this and need to be added in these - are Norman 1987 and cooke 2005
Full size table

Of the databased articles (n = 68), 17 (25%) studies were mixed methods, 23 (34%) were qualitative and 28 (41%) were quantitative. The majority (n = 49, 72%) were categorised as descriptive studies, 17 were intervention studies (25%) and two were measurement studies (3%). Of the non-databased articles (n = 84), the majority (n = 46, 55%) were commentaries, one was a case study (1%) and the remainder were program descriptions (n = 37, 44%).

  1. ii.

    Study design

Of the 18 data based studies that were interventions, only two were controlled studies [74, 107] and one was a protocol of a controlled trial [142]. Nine studies were post-intervention, cross sectional assessments without control arms [34,35,36,37,38,39,40,41,42], five were prospective or retrospective cohorts, pre-post studies, or repeat cross sectional assessments [54,55,56,57,58].

  1. iii.

    Study setting and participants

Of the 68 databased articles, 29 studies were conducted either in hospital(s) or a department within a hospital, seven were in health care settings more broadly, four were in community health care settings, 16 explored networks (usually a collaboration of academic/health care partnerships and institutions), 11 were in primary care, and one in a university. There were a range of participants in these studies including Aboriginal Health Workers, nurses, midwives, physiotherapists, primary care practitioners, pharmacists, allied health practitioners, community health care workers and practitioners, clinicians and administrators within particular health systems or context (i.e. NHS, CLARHC), counsellors, ICU staff, multidisciplinary primary health care team and medical students.

Qualitative synthesis of strategies

Table 3 outlines the strategies identified in all studies that described an intervention (both databased and non-databased) and examples of suggested/actual application in practice. We identified and described eight categories of strategies: dual skilled team/staff, resources or physical infrastructures, incentive, leadership support of research, education/training, networks, forming partnerships or collaborations and overall leadership for entity.

Table 3 Qualitative synthesis of research engagement strategies outlined in all included studies
Full size table

Discussion

This scoping review sought to provide an overview of the existing research exploring strategies that have been suggested to improve provider and health care organisation’s engagement with research. This review found that over half of publications describing research engagement did not formally present any new data (non-databased and opinion pieces). Of the databased publications (those that formally presented new data), the majority were descriptive in nature largely providing evidence of the extent of the behaviour and association between strategies and research engagement. Our study found only two controlled trials evaluating the impact of research engagement strategies [74, 107]. Similarly, the review by Gagliardi et al. in 2015, also found only 13 studies, none were controlled studies, describing how iKT approaches have been operationalised and used to improve collaboration between researchers and decision makers in the research process [15]. Such findings are consistent with that of an emerging field, however, are of concern given significant resource investment by governments internationally to establish structures and implement strategies to build research leadership and capacity in health care organisations. For example, the NIH Roadmap allocated an initial $125 million in 2004, with planned increases to $2 billion in the later stages of implementation [186] where funds are distributed across developing new innovations, developing new models for research and increasing research translation.

Encouragingly, the percentage of databased and intervention studies appears to be increasing over time although the overall number is still small. Using a FA, we identified and described strategies that were suggested to be useful, or had been previously applied to increase clinician and health care organisations engagement with research, with the intent of providing a list of strategies that could be applied and evaluated in future studies. The research engagement strategies were described across eight categories targeting clinician, team, organisation and supra organisation factors [71]. Such strategies seek to target the broader determinant described by Gagliardi et al. and are consistent with that outlined in the research co-production literature [187], which suggests that interventions need to move beyond targeting just knowledge and skills of clinicians. Findings from this review also add to research capacity building frameworks for practitioners by describing the broader collaborative (e.g. networks), structural or workforce arrangements (e.g. clinician researchers) that could be put in place to support clinical research leadership [16]. Further intervention research is warranted targeting these eight distinct categories to better understand the impact of such strategies on increasing clinician research engagement. Additionally, a systematic review of the descriptive literature may also provide additional insights into the association of such strategies with engagement outcomes.

Our review highlights a mismatch between investment in research engagement strategies and the available evidence to support such strategies. While initiatives like the NIHR CLARHC provided significant opportunity and funding to embed a broad range of strategies to facilitate research co-production and clinician-led research, the impact these structures have on research engagement have yet to be examined [188]. Further, to our knowledge, there have been no controlled evaluations of multi-level formalised programs such as CLAHRC [188]. This may be due to the challenges with defining important outcomes of multi-level initiatives that account for the individualised approaches taken by different sites to contextualise the intervention. These non-controlled evaluations provide rich and important data to understand the factors [189] that facilitate implementation and the context in which these initiatives are delivered. However evidence from these non-controlled evaluations needs to be coupled with rigorous comparative evaluations to provide essential evidence for decision-making and to justify continued investments.

Strengths/limitation

This study used high quality, systematic processes to ensure that a broad range of studies examining research engagement were included. At least two screeners were included in each of the review processes. The use of structured qualitative processes to generate a list of strategies allowed for a broad examination of the types of strategies previously discussed or applied in research and practice, and enabled a collaborative analyses by researchers with varying degrees of qualitative research experience.

While a comprehensive search strategy consisting of a database search, search of the grey literature and consultations with experts in the field was used, it is possible that studies were missed due to inconsistent terminology and the rapidly progressing evidence-base. We excluded non-English studies and as such could have missed potentially relevant studies published in other languages. Nevertheless, this study to our knowledge provides a comprehensive overview of the characteristics of studies examining research engagement using high quality processes. This scoping review highlights that future empirical research is needed to identify the impact of such proposed strategies on health services, practice and patient outcomes.

Conclusions

This scoping review for the first time provides a list of research engagement strategies that have been proposed to be useful to increase clinicians and health care organisations collaboration in research activity. The majority of the evidence however has been descriptive in nature, providing limited empirical evidence to support the efficacy of research engagement strategies. There is a need for future research to progress beyond descriptive research to methodologically rigorous intervention research, to provide the evidence needed to inform decision making.