Researcher practitioner engagement in health research: The development of a new concept.

The engagement of frontline practitioners in the production of research-derived knowledge is often advocated. Doing so can address perceived gaps between what is known from research and what happens in clinical practice. Engagement practices span a continuum, from co-production approaches underpinned by principles of equality and power sharing to those which can minimalize practitioners' contributions to the knowledge production process. We observed a conceptual gap in published healthcare literature that labels or defines practitioners' meaningful contribution to the research process. We, therefore, aimed to develop the concept of "Researcher Practitioner Engagement" in the context of academically initiated healthcare research in the professions of nursing, midwifery, occupational therapy, physiotherapy, and speech and language therapy. Guided by Schwartz-Barcott et al.'s hybrid model of concept development, published examples were analyzed to establish the attributes, antecedents, and consequences of this type of engagement. Academic researchers (n = 17) and frontline practitioners (n = 8) with relevant experience took part in online focus groups to confirm, eliminate, or elaborate on these proposed concept components. Combined analysis of theoretical and focus group data showed that the essence of this form of engagement is that practitioners' clinical knowledge is valued from a study's formative stages. The practitioner's clinical perspectives inform problem-solving and decision-making in study activities and enhance the professional and practice relevance of a study. The conceptual model produced from the study findings forms a basis to guide engagement practices, future concept testing, and empirical evaluation of engagement practices.

is clear (Dimova et al., 2018;Marjanovic et al., 2019). Their skillset, personal characteristics, and existing relationships can support patients in the process of choosing to take part (Cronin et al., 2019;Lavender et al., 2019;Mann et al., 2014). Practitioners' clinical roles also make them well placed to deliver study interventions as part of routine care (Boase et al., 2012;Stockwell-Smith et al., 2015).
There is a risk, however, that when practitioners execute study protocol activities, a form of engagement known as a hired hand approach can be adopted (Daniels et al., 2020;Roth, 1966; Table 1).
Hired hand research is experienced by those who follow a pre-formed plan laid out by the researcher (Roth, 1966). Examples demonstrate how, in such cases, practitioners are offered little opportunity to influence a study. As a result, their behaviors can affect a study's outcome, with the potential to threaten the quality of the data collected (Dyson & Dyson, 2014;Poat et al., 2003).
Conversely, practitioners' engagement in research can be highly collaborative. A range of theoretical propositions such as participatory methodologies, Mode 2 knowledge production, engaged scholarship, and integrated knowledge translation have at their core a high level of cooperation between those who produce research and beneficiaries. The phrase "co-production of knowledge" is consistently associated with collaborative approaches. This term portrays a process through which researchers and research users undertake a study together (Antonacopoulou, 2010;Armstrong & Alsop, 2010). Co-productive approaches are driven by the need to engage with those likely to act on the knowledge that is generated (Nutley, 2010) with the specific goal of increasing the application of research through relevant, better quality studies (Bowen & Graham, 2013). An approach underpinned by these engagement principles (Table 1) demonstrates a clear endeavor to engage research users in all or most study activities, coupled with equality and power sharing across the research process (Beckett et al., 2018).
We scoped peer-reviewed publications for literature that empirically evaluated or described examples of frontline practitioner engagement by academic researchers (Daniels et al., 2020). The type of engagement observed often did not fully align with the defining characteristics of the engagement paradigm. Practitioners were more likely engaged in only some research activities, usually T A B L E 1 Comparison of the characteristics of the "hired hand" approach and the engagement paradigm Hired hand approach (Roth, 1966) Engagement paradigm (Bowen & Graham, 2013) Who Who Hired hand: those assigned a task within a study by the researcher Knowledge user: those who will act on the knowledge generated by a study

Why Why
Achieve researcher's goals Co-production of knowledge Activities Activities Assigned tasks (e.g., participant recruitment or data collection) Researchers and knowledge user collaboratively make decisions on: | 535 recruitment, data collection, and/or intervention delivery. In these cases, there was, however, evidence to suggest that the practitioner's role had resulted in positive effects for the study, clinical practice, and/or practitioner development (Boase et al., 2012;Bullen et al., 2014;Campbell et al., 2015;Eriksson et al., 2013).
Notably, our review found inconsistency and variation in the terms used by authors to refer to the form of engagement we had observed. It could be argued that this activity aligns somewhat with the concept of stakeholder engagement, but as a broad term is not specific to or often inclusive of frontline practitioners (Camden et al., 2015;Concannon et al., 2012). Engagement with all user groups is advocated within a research study to address different realities and perspectives as each group brings different motivations, expectations, and cognitive and emotional perspectives to the research process (Rycroft-Malone et al., 2016). However, strategies are required that specifically address variations in the engagement needs of each user group (Henderson et al., 2014). We, therefore, identified the need to develop a theoretical concept specific to this form of practitioner engagement by researchers based in academic institutions. Defining a form of engagement that converges around one specific term could open conversations and address current inconsistencies and limitations in the reporting of engagement practices (Daniels et al., 2020).

| Aim and objectives
The aim was to develop the concept of "Researcher Practitioner Engagement" (RPE) in the context of academically initiated healthcare research in the professions of nursing, midwifery, occupational therapy, physiotherapy, and speech and language therapy. Specifically, the objectives were to establish what constitutes the concept by identifying the attributes, antecedents, and consequences to both define and delineate it from other concepts, determine if the concept is deemed necessary, and confirm suitability of the concept label.

| Study design
RPE is poorly developed, poorly explained, and has a lack of defined parameters, therefore, is not easily discernible in the literature . The immaturity of the concept necessitates an approach that is not reliant solely on theory, but which enables experiential data to form part of the concept development process. Therefore, qualitative methods which allow for an inductive approach were used . The hybrid model of concept development (Schwartz-Barcott et al., 2000) was adapted to optimize the rigor and usefulness of the results. In a three-phase approach, theoretical strategies and qualitative methods were combined to produce outcomes based on both literature and empirical data developed from actual cases .

| Theoretical phase
Sources used in this phase were 10 instances of the observed phenomenon which had been identified in peer-reviewed publications via a scoping review conducted in October 2017 (Daniels et al., 2020) and one instance retrieved by repeating the search 6 months later (March 2018). These instances were detailed in empirical evaluations of practitioner engagement by academic researchers in the research process (n = 8) and in descriptive papers designed specifically to report an engagement example (n = 3). Definitions of the related concepts stakeholder engagement (Concannon et al., 2014;Deverka et al., 2012), practitioner researcher engagement (Brown et al., 2001(Brown et al., , 2003, and engagement in healthcare (Norris et al., 2017) were identified during the literature search. As these sources referred to defining elements of practitioner engagement in research they were also used.
Sources were transferred to and managed in NVIVO® (version 11, 2015). Using qualitative content analysis (Elo & Kyngas, 2008;Mayring, 2014), factors required for RPE to occur (attributes), conditions necessary before RPE can take place (antecedents), and outcomes of RPE (consequences) were extracted. Within each category, subcategories were inductively generated by grouping similar or related components and naming each with a representative label (Elo & Kyngas, 2008). This process was iterative as subcategories were revisited and recategorized through continual reflection and abductive inference (Krippendorff, 2013) and continued until all evident conceptual components were identified.

| Fieldwork phase
Using focus groups, perspectives of academic researchers and frontline practitioners with engagement experience were used to confirm, refine, expand and/or exclude the tentative attributes, antecedents, and consequences inferred from the theoretical phase.  (Mayring, 2014), patterns in reasons for confirmation, refinements, or elaborations of each concept component were identified.
Participants' views on the necessity of the concept and concept label were analyzed and reasons categorized.
T A B L E 2 Inclusion criteria for participants in the fieldwork phase of the concept development

Inclusion criteria
Academic researchers or doctoral researchers based in faculty/college of health-related subject areas within higher education institutions in the United Kingdom Frontline practitioners (nursing, midwifery, occupational therapy, physiotherapy, speech and language therapy) delivering care to service users in a healthcare context Principal investigator of at least one health-related research study completed within the past 3 years (concerning nursing, midwifery or occupational therapy, physiotherapy, speech and language therapy practice) Engagement by an academic researcher from a University setting in at least one health-related research study (other than as a participant) within the past 3 years Self-reported experience of engagement of practitioner(s) in a role other than as a study participant in at least one research project in the past 3 years

Exclusion criteria
Employed solely within a health setting In a role with formal research responsibilities (e.g., clinical research nurse, clinical academic, research therapist) Solely employed with a hybrid or cross-organizational initiative or system specifically funded to support collaborative practices across academic and health organizations T A B L E 3 Outcome of theoretical phase of concept development

| Analytical phase
The purpose of this final phase was to integrate the literature and empirical data (Schwartz-Barcott et al., 2002). This involved moving iteratively between focus group data and returning to data from the theoretical phase to ensure sound representation of each component before establishing the concept definition.

| Rigor
To establish validity through confirmation and enhance understanding of the concept (methodological triangulation), four academic researchers (Focus group R5) were not exposed to the outcome of the theoretical phase and instead were asked to identify the attributes, antecedents, and consequences of the concept solely from their experiences. This focus group was facilitated by a researcher (P.G.) who had not been exposed to the outcome of the theoretical phase. Triangulated data were mapped to the theoretical phase outcome to identify convergences and additional concept components, helping to establish validity both through confirmation and by enhancing understanding of the concept through completeness (Breitmayer et al., 1993;Risjord et al., 2009). Recruitment challenges prevented triangulation with practitioner participants. As academic researchers, and therefore "insiders" (Finefter-Rosenbluh, 2017) reflexivity was essential and ensured through critical selfreflection of our positionality (Berger, 2015), identifying any potential influences on the data collection and analysis and monitoring any potential effects through an audit trail of interpretations maintained in a journal. The journal was added throughout to the theoretical and fieldwork phases to record researcher interpretations and was a key tool in the analytical process. Member checking of key discussion points with all participants highlighted no disagreements with accuracy.

| Ethical considerations
Ethical approval was obtained from the Institute of Nursing and Health Sciences Research Governance Filter committee. Key considerations of study involvement, handling and privacy of data, and withdrawal procedures were communicated during the recruitment phase. Informed written consent included an agreement for audio and visual recordings of discussions.

| FINDINGS
Five attributes, five antecedents, and three consequences were identified in the theoretical phase (Table 3).
Seventeen researchers and eight practitioners met the study criteria and were available to take part in eight focus groups con-

| Antecedents
Participants' views stemmed mainly from barriers and facilitators experience which gave insight into the conditions necessary for RPE.

| Establishing the need for this concept
In the main, participants agreed that the concept of RPE is necessary.
Reasons to support this were categorized as (a) to improve engagement practices and (b) to legitimize this form of engagement. Comparisons were drawn with Patient and Public Involvement, citing the positive consequences that formally establishing and building a culture around this subgroup of research users had realized. Despite an overall sense that the concept would be useful to guide successful engagement practices and overcome potential barriers, there were some reservations.
Engagement was viewed as integral to the work of one researcher who did not believe RPE needed to be extrapolated as a separate entity. But, it was also felt that engagement does not happen intuitively. Improving understanding could prevent researchers taking engagement for granted and highlight what needs to be addressed to ensure engagement happens in a meaningful way. Researchers with reservations, however, did recognize the benefits of thinking carefully about a practitioner's role as opposed to merely demonstrating clinical input in funding applications.
This was echoed by a practitioner who voiced the need for a culture where approval committees and funding bodies require explicit evidence of RPE. A definition was also felt important to facilitate consistency in engagement practices and language used, allowing for comparatives to be made, impact of engagement to be measured and an evidence base developed. Most agreed that the label "Researcher Practitioner Engagement" was representative of the concept and its components. Alternatives such as "partnership" were proposed but challenged as being overly formal whereas engagement was thought to represent the concept's fluidity.

| Outcome of the analytical phase
The experiential lens of participants enabled the concept components to be refined to their most salient elements and provide a sound representation of the concept of RPE. No element of the concept proposed in the theoretical phase remained unchanged; most components were refined or removed and one component initially proposed as a consequence became a defining attribute. The final concept components detailed in Table 5 were used to propose a tentative definition: "Researcher Practitioner Engagement is a mutually beneficial process, through which practitioners are engaged by researchers to actively contribute to the production of research-derived knowledge which is meaningful to their practice. Practitioners' clinical perspectives, skills, and/or knowledge influence a study from its formative stages and, through open dialogue, are used to problem solve and inform decisionmaking in relevant study activities to optimize the clinical relevance of the study and its outcomes." The outcome of the analytical phase was used to devise a conceptual model to diagrammatically represent relationships between the concept components and to optimize its usefulness in guiding RPE in healthcare research (Figure 1).

| DISCUSSION
The concept of RPE responds to the concern that opportunities for practitioner engagement in research need to be realized (Marjanovic et al., 2019;McCormack, 2011;Pentland et al., 2011). It addresses the view that those who provide clinical services should be included in studies so their skills and strengths are capitalized on to enhance study tasks (Cronin et al., 2019;Nelson et al., 2007). RPE's central intentionality is to ensure a practitioner's clinical perspectives influence a study and its outcomes. The value placed on practitioners' experiential knowledge within this new concept mirrors a central component of the engagement paradigm. The key to this existing paradigm, however, is that research users and producers collaboratively make decisions in relation to all or most study activities (Bowen & Graham, 2013). But, from the perspectives of both researchers and practitioners within this study, this was deemed neither necessary nor feasible. The notion of shared decision-making was contested with researchers being clear that a study is ultimately their responsibility, a sentiment with which some practitioners agreed. Practitioners expressed the need to feel their clinical perspectives are of equal value to the scientific perspectives of researchers generally, and used to influence the research process, particularly at the formative stage. This was endorsed by practitioners as more feasible in light of other clinical priorities than alternatives that require them to take on greater responsibility and commitment. Early engagement with clinicians is essential to understand how the study can be integrated into current clinical workflow and the adaptations necessary to ensure a study is acceptable to the clinicians concerned (Topazian et al., 2016;Weinfurt et al., 2017). Although the ideal of co-production of knowledge is postulated, few reported examples of practitioner engagement by academic researchers conform to the characteristics of this approach (Daniels et al., 2020). Evidence to demonstrate the impact of co-production on the relevance and utility of a study is sparse, outside of participatory action research approaches. It is, therefore, difficult to create a strong argument that supports the ideal of engaging frontline practitioners in all or most study activities. This is not of course to say that this ideal should not be strived for. However, the challenges of doing so must be acknowledged, and feasible ways of achieving collaborative knowledge production recognized (Rycroft-Malone et al., 2016).
The researcher-initiated agency of this concept could be seen to contradict the egalitarian, bottom-up approach of participatory approaches, in which practitioner-initiated studies are advocated as most likely to produce relevant research (Blevins et al., 2010). Power imbalances could also pose a challenge to the success of collaboration (Brown et al., 2003;Rycroft-Malone et al., 2016). However, engagement in protocol design is considered a defining distinction of a collaborative approach (Nelson et al., 2007). Therefore, the requirement for practitioner engagement in devising the study protocol could contribute to flattening knowledge hierarchies. Tangible recognition of a practitioner's perspectives in the study design could eliminate practitioner frustration when this does not occur (Blevins et al., 2010) and provide opportunity to ensure aspects of the study design are acceptable to all parties (Newington & Metcalfe, 2014).

| Implications for practice
This new concept addresses, in part, the variable and inconsistent terminology used to describe this engagement activity, which has led to challenges when carrying out reviews in the engagement field (Concannon et al., 2014;Fransman, 2018;Malterud & Elvbakken, 2019).
It adds to the somewhat limited theory available to guide engagement practices to realize outcomes that could positively impact the research-practice gap. Defining components are mirrored in related work conducted since, which also identifies key considerations when involving healthcare practitioners in the research process (Laustsen et al., 2020). Laustsen et al's (2020) adaptation process model similarly emphasizes healthcare practitioners' contentment at being led by the researcher but a clear desire to advocate for a project's applicability, which subsequently strengthened their practice.
By proposing this new concept, it is anticipated that RPE will be recognized, begin conversations, generate new examples, and the T A B L E 5 Outcome of analytical stage: The components of the concept "Researcher Practitioner Engagement"

Attributes Antecedents Consequences
(1) Engagement in study activities varies but always occurs in protocol design and dissemination stages (1) Common vested interest in a study topic and its outcomes (1) Improves clinical relevance of a study and its outcomes (2) Practitioners' perspectives, skills and/or knowledge influence the research process from the formative stages (2) Initiation and forming of a collaborative relationship (2) Practice development We hypothesize that RPE could prevent engagement practices from adopting a marginalized, hired hand approach, which has the potential to threaten the feasibility and quality of the research process and a study's outcomes (Dyson & Dyson, 2014). Evaluations of recruitment practices within clinical trials have shown that when the understanding of a study is not in place, clinicians negatively perceive the study's relevance to their clinical practice, which, therefore, affects who is recruited (Ziebland et al., 2007). Those who provide clinical services should, therefore, be included in the planning of studies as a strategy to reduce gatekeeping behaviors (Cronin et al., 2019). The clinical skills and strengths of practitioners can then be capitalized on to enhance study tasks (Morrison-Beedy et al., 2001;Nelson et al., 2007).
Behaviors such as study referral are considered more likely if clinicians feel a sense of ownership, hold positive views of the intervention being evaluated , and understand the methodology being used (Lamb et al., 2016).
Increasing the need to demonstrate a study's impact means it is imperative researchers ensure findings can be utilized in practice. This necessitates a balance between rigor and relevance (Rothmore, 2018).
Considering these as discrete requirements could create a barrier to knowledge derived from research fulfilling its intended function of providing evidence to inform healthcare practices and optimize patient care.
Collaboration between researchers and practitioners is, therefore, essential to inject realism into study design (Pickler & Kearney, 2018) and represent the "real clinical world" (Patterson et al., 2010). Consequently, consideration of research relevance (i.e., external, social, and ecological validity) which is equitable to the consideration given to robustness and internal validity in a study's design is advocated (Backus & Jones, 2013).
Perhaps it is time to revisit Roth's (1966) assertion that critical appraisal of how knowledge has been produced should include evaluating if a hired hand approach has been adopted and subsequent impact. Strategies adopted to assure clinical relevance in study design should be called upon to be transparent in reporting as a matter of course.

| Methodological considerations
Perspectives of researchers and practitioners with engagement experience were used to confirm, refine, expand, and/or exclude the  Council, 2018;Nursing and Midwifery Council, 2015). By proposing this concept, we hope to open discussion on its potential for helping to develop a culture that works toward achieving co-productive ideals and prevent a hired hand approach that marginalizes the contribution practitioners can make to the research process. By fostering a culture supporting co-productive ideals, RPE may, thereby, optimize research outcomes and their utilization in practice.

This project was funded by the Department for Education and
Learning (DEL), Northern Ireland.