Enabling Visibility of the Clinician-scientists’ Knowledge Broker role in Linking Research with Clinical Practice: A Participatory Design Research in the Dutch Nursing-home Sector

Background: Twelve clinician-scientists were employed in a Dutch academic network, which is a collaboration between fteen nursing-homes and an academic medical research institute. The clinician-scientists were tasked with linking research and clinical practice by catalysing both care-informed research and evidence-informed implementation initiatives. The clinician-scientists and their manager experienced diculties in clearly dening the knowledge broker role of the clinician-scientists, a diculty also reported in literature. They found no tools and methods suitable for make their knowledge broker role visible. Clarifying role expectations and accountability for funding these knowledge broker positions was dicult. They aimed to design a theory-informed performance appraisal tool that allowed clinician-scientists to explicate and develop their knowledge broker role in collaboration with management. Methods: A participatory design research was conducted over a 21 month period with a design group consisting of an external independent researcher, clinician-scientists and their managers from within the academic network. Results: A tool (the SP-tool) was developed in MS Excel. This allowed clinician-scientists to log their knowledge broker activities as distinct from their clinical work and research related activities. The tool contributed to their ability to make their knowledge broker role visible to themselves and their stakeholders. The theoretic contribution of the design research is a conceptual model of professionalisation of the clinician scientists knowledge broker role. This model presents the relationship between work visibility and the clarication of functions of the clinician-scientist’s knowledge broker role. In the professionalisation of knowledge-intensive work, visibility contributes to the denition of CS broker functions, which is an element necessary for the professionalisation of an occupation. Conclusions: The CSs knowledge broker role is a knowledge-intensive role and work-tasks associated with this role are not automatically visible. The SP-tool contributes to creating work visibility of the clinician-scientists’ knowledge broker role. This in turn could contribute to the professionalisation of this role, which is not well described in literature at the day-to-day professional level.


Background
The value of clinician-scientists (CSs) in linking the often disparate contexts of research and clinical practice, by virtue of having professional jurisdiction in both, is uncontested in the discourse of translational research (1). As knowledge brokers, CSs have the advantage of habitual, rst-hand experience of clinical work and research processes. This facilitates strategic networking (2) and the involvement of stakeholders (3,4) in designing and executing broker activities that link research and practice (5). Operational involvement in both research and clinical practice, facilitates unique insights into the goals, priorities and organisational logics of both contexts (6). CSs are in a position to display sensitivity and responsiveness to both contexts (4). This strengthens CSs ability to design tenable broker activities that balance scienti c and economic interests and consequently have a higher potential to appeal to both scientists and care providers (5).
The knowledge broker role of CSs is distinct from the clinical-and research roles (5) and it requires competencies additional to those required to execute clinical-and research tasks (5). The knowledge broker role comprises activities related to knowledge management, linkage and exchange, and capacity building (7). Successful knowledge brokering by CSs leads to an increased volume of clinically relevant research results (2,8,9) and increased evidence application in practice (10,11). The nature of the connection brokered by CSs is ideally bilateral and dynamic in nature, whereby the research context and the clinical context inform each other (6).
Despite the espoused value of the CS as a knowledge broker, little published information exists about the exact nature of this role at the day-to-day professional level. The connection between research and practice is frequently assumed to occur by virtue of the CSs professional jurisdiction in both settings (5).
The knowledge broker role of CSs is not yet concrete enough to constitute a professionalisable work package that lends itself to the establishment of exclusive professional jurisdiction (1,12). CSs occupy an ambiguous intermediary position between research and practice(13), they struggle with professional identity issues and workload across disparate roles (1,13). Their research and clinical activities are visible, easily quanti ed and thus contributing to the establishment of subject matter expertise or specialisation. Their broker activities, however, remain unclear and potentially unseen, whilst these activities consume a substantial portion of time. Some broker activities are viewed as logistic and generic tasks that do not bestow professional expertise and hence do not require professional development effort in their own right (14), for example implementation facilitation (4). CSs unsurprisingly experience workload di culties and CS numbers are declining (13). The aim of this research was to identify a manner in which the knowledge broker role of CSs could become more visible and tangible at the day-to-day professional level. Higher visibility and understanding of this role might lead to its tenable development and professionalisation.

Context of this research
This research was conducted in a Dutch academic network: a collaboration between fteen nursinghomes and an academic medical research institute. As part of the strategy to link research and clinical practice, the academic network employed twelve master-educated CSs in 2018 and 2019. The CSs were tasked with catalysing both care-informed research and evidence-informed implementation initiatives. The managers of the CSs voiced concerns about their limited ability to demonstrate accountability for funding these broker positions. The CSs reported insecurities about role-expectations and di culties in making their broker role visible. The di culty in clearly de ning the CSs broker role presented a practical challenge in this academic network. The managers and CSs were unable to nd a tool suitable for making the CS broker role visible. In response to this practical di culty, we conducted a participatory design research aimed at developing a practical method of making the broker role of the CSs visible.

Methods
Design research is a genre of research that is collaborative and suitable for design and construction of tools that are required to solve practical problems (15). It contributes to existing theory (15), in this case, the theory on the visibility and professionalisation of the CS broker role. Design research attempts to balance research rigour with practical relevance.
We invited all CSs and managers from within the academic network to form a design-group together with an external independent researcher (MB), with the aim of designing a tool that allows CSs to make their broker role visible. We implemented the three phases of design research (15). Phase 1, the orientation phase, consisted of a literature review, a context analysis and a needs analysis to explore the needs of the academic network, in order to draft a design requirements and a design proposition (15). Phase 2 consisted of the cyclical process of design, construction and evaluation of a broker-activity logging tool, that ful lled the design requirements drafted in phase 1. Phase 3 entailed evaluation of the nal tool designed. These three phases and the concomitant data collection activities are summarised in Fig. 1.
The literature review in phase 1 was conducted from May to August 2017 according to the realist review method and was published separately (5). The literature review served as a starting point from which the design-group worked collaboratively during the remainder of the project, which entailed six-weekly meetings from July 2018 to July 2019. The nal evaluation extended to March 2020.
In Phases 1 and 2, data were collected during design-group meetings, individual interviews with CSs, participant observation of supervision meetings between CSs and their manager, questionnaires, walkthroughs, micro-evaluation of elements of the designed tool and try-outs of the tool. Data were in the form of minutes of meetings, sociograms, notes in the researchers log book and member checked notes taken during interviews and conversations. Key assertions by participants were documented verbatim and member checked. In phase 3 an anonymous online questionnaire was distributed to CSs. They rated their perception of the tool's effect on their ability to make their broker role visible, on a seven-point Likert scale ranging from 'a lot worse' to 'a lot better', compared to a situation without the tool. CSs also indicated on two multiple choice questions how they planned to use the tool. An individual discussion and re ection was conducted with the manager after the nal try-out of the tool.
We analysed data collected in phase 1 using framework analysis (16), which is suitable for research that develops new plans and actions. The independent researcher analyzed the raw textual data following the steps of coding, indexing, charting, mapping and interpretation(16) using a framework in Microsoft Excel.
The codes comprised the inner-and outer context factors necessary for effective CS brokerage as identi ed in the literature review (5). A critical friend (AP) oversaw parts of the analysis process to enhance credibility.
During phase 2 (design and construction), data were collaboratively translated into changes required for the prototypes of the tool during design-group meetings. The CSs and the manager made practical suggestions and the independent research introduced a theoretic perspective. The answers from the online evaluation questionnaire in phase 3 were numerically tallied and reported in raw numbers.

Results
The design-group consisted of eleven of the twelve CSs and their manager. One CS did not consent to participation. All participants gave written informed consent to participate in the research. The professional backgrounds of the CSs were in nursing, physiotherapy, speech and language therapy, sportand-movement therapy and psychology. One CS ful lled a part-time management role within the network. Nine of the eleven CSs were clinical professionals within care facilities, who received funding by the academic network to dedicate one day per week to the CS broker role. Two were researchers employed by the academic research institute, who received funding to spend one day per week as a CS within an assigned care facility.
The practical result of this design research was a tool in ful lment of the design proposition: 'to design a performance appraisal tool that makes the broker role of CSs visible'. We called the nal product, the Science-Practitioner tool (SP-tool) (see Fig. 2). The SP-tool was the result of three design cycles and made it possible for CSs to document their broker activities exibly. The six collaboratively formulated design requirements for this tool were as follows: 1 -data entry is not time consuming; 2 -work in progress and work completed are explicated; 3 -an individual pro le of the CS is presented; 4organizational barriers and facilitators to goal achievement are presented; 5 -broker activities are linked to organization priorities; and 6 -the manner in which the CS is a catalyst in linking research and practice is shown.
In the rst design cycle, a narrative report format with pre-determined headings and sub-headings was designed and used. The report formed the basis for discussion during the CSs performance appraisal with their manager. This format was not deemed suitable by CSs and managers alike as it was static in nature. The managers could not ascertain common elements between reports in order to gain an understanding of the broker role independent of context.. One manager (participant 1) said, "We want to be able to assess whether the CSs are doing enough but we don't know what 'enough' is." The CSs found that documenting their goals for the broker role on a Goal Attainment Scale(17) was too rigid and too speci c. It did not assist in showcasing exibility and sensitivity to the context. "I need to show the activities I do within the broader goals of the CS function" (participant 10). CSs felt it important to showcase all the requests they received and their subsequent decisions in prioritizing certain requests over others. "I receive a tsunami of questions but many are not suitable for a CS, more so for a researcher. I want to communicate more about my (CS) function." (participant 8). The report did not make the dynamic nature of the CSs network activity and social capital visible. Respondent 11 said, "I link people in my network with each other and great things happen but this is not visible anywhere".
Elements of the report that were seen as positive were that CSs could present their professional pro le, the organizational barriers and facilitators to goal achievement and the categorization of the broker activities they executed in a theory informed framework of knowledge broker roles(7).
To address the di culties, the second and third design prototype were MS Excel spreadsheets, which allowed all requests received by CSs to be documented, not only those that were accepted and operationalized by CSs. For each request logged, the CS could input data from dropdown lists about the nature of the request, the network partner involved in submitting the request, the complexity of work involved in operationalizing this, the relationship to the strategic goals of the organization and the nature of the broker activity required to address the request. Table 1 gives an example of entries made by CSs. The goals of the CS broker role were formulated more broadly and were able to accommodate a broad spectrum of requests from stakeholders. In contrast with the rst prototype, the reporting structure in MS Excel secured an element of uniformity in reporting. This assisted managers in their understanding of the broker role. It allowed visibility of the conceptualisation which the stakeholders (network) have of the CS broker role, as seen in the type of requests for engagement. It also presents the conceptualisation which the CS has of their own role by showing which activities he/she prioritises and why. Seven of the eleven CSs completed the online questions about the usefulness of the SP-tool. They anticipated it as useful in making their tasks visible to stakeholders and improving their own understanding of the broker role. Five indicated that owing to the SP-tool, their ability to explicate their broker role was 'much better' and two indicated that it was 'somewhat better'. All (n = 7) intended using the SP-tool for accountability purposes towards management, ve intended using it to gain insight into questions from the clinical practice environment and as a basis for discussion with colleagues and funders. Four intended using it to plan and prioritise their work in the CS broker role. The manager implemented the SP-tool for use by all CSs in the next performance appraisal meeting of each CS.

Discussion
This design research has been useful in developing a practical theory-informed tool (the SP-tool), which allows individual CSs to make their knowledge broker role in a nursing-home context visible. Visibility of knowledge-intensive work is not automatic but intentionally constructed (18). The SP-tool assists with intentional construction as it allows CSs to make broker activities, that belong neither exclusively to the jurisdiction of research nor that of clinical practice, structurally visible during performance appraisal.
Strategic construction of work visibility based on active engagement in an organisation, contributes to the professionalisation of knowledge-intensive occupations (19). This dynamic form of professionalisation stands in contrast to that of traditional healthcare professions, in which professionals are socialised to work according to a clearly de ned and visible scope of practice. Professionalisation through engagement in a dynamic context, where the expectations and views of stakeholders inform the development of an occupational role is a process that might be new an unknown to CSs who stem from an academic health professions background. To illustrate this process we contribute a conceptual model based on the ndings of this design research. The model integrates work visibility with theoretic tenets of professionalisation for the CS broker role. A core element of professionalism is the existence of clearly de ned work functions (20) and a shared understanding of these functions between the professional and the service-user. In our model (see Fig. 3), engagement with employers and service users, is a means to de ning professional remit (19). A de nition of functions constituting the occupational role of the CS is postulated through dynamic interaction in the work context. The de nition of functions is not pre-determined by a state regulated monopoly such as is the case for medical professionals. We would like to propose this conceptual model for professionalisation of the CS broker role. Figure 3 presents the conceptual model.
Professionalisation is said to improve the reputation of a profession (21). Interestingly, CSs are generally viewed positively and deemed important(5), despite their di culties with role clarity.
In addition to its contribution to professionalisation, the SP-tool can be viewed as a boundary object (22), contributing to a common understanding between the CS and their manager. A collective understanding of professional role contributes to a collective identity of a profession (20) and might contribute to overcoming professional identity di culties associated with the CS broker role.
A limitation of design research is its context speci c nature and very limited generalisability. However, the clearly de ned design proposition and requirements in this research might be of interest to CSs and managers in other contexts. Furthermore, the SP-tool and the proposed conceptual framework could be of interest in future more fundamental research into the day-to-day professional level of CSs. The small number of participants is a limitation, however, in the academic network all but one CS participated. Their contribution signi cantly enhanced the practical applicability of the SP-tool.

Conclusion
The CSs broker role is a knowledge-intensive role and work-tasks associated with this role are not automatically visible. The SP-tool contributes to creating work visibility of the CSs knowledge broker role, which might assist in the clari cation of CS broker functions and contribute to professionalisation.

List Of Abbreviations
CS: Clinician Scientist