The roles, activities and impacts of middle managers who function as knowledge brokers to improve care delivery and outcomes in healthcare organizations: a critical interpretive synthesis

We used a critical interpretive synthesis (CIS) to systematically review the complex body of literature on MM KBs. This included qualitative, quantitative, and theoretical papers. CIS offers an iterative, dynamic, recursive, and reflexive approach to qualitative synthesis. CIS was well-suited to review the MM KB literature than traditional systematic review methods because it integrates findings from diverse studies into a single, coherent framework based on new theoretical insights and interpretations [26, 27]. A key feature that distinguishes CIS from other approaches to interpretive synthesis is the critical nature of the analysis that questions the way studies conceptualize and construct the topic under study and uses this as the basis for developing synthesizing arguments [26]. We ensured rigor by complying with the Enhancing Transparency in Reporting the Synthesis of Qualitative Research (ENTREQ) criteria (Additional file 1) and other criteria of trustworthiness [28, 29]. We did not register a protocol for this review.

With a medical librarian, we developed a search strategy (Additional file 2) that complied with the evidence-based checklist for peer review of electronic search strategies [30]. We included Medical Subject Headings and keywords that captured the concepts of MMs (e.g., nurse administrator, manager), explicit or non-explicit KB roles (e.g., diffusion of innovation, dissemination, broker, and facilitator), evidence-based practice (e.g., knowledge, evidence) and setting (e.g., hospital, healthcare, or health care). We searched MEDLINE, CINAHL, Social Sciences Abstracts, ABI/INFORM, EMBASE, PubMed, PsycINFO, ERIC, and the Cochrane Library from January 1, 2001, to August 14, 2020. We searched from 2001 onward because the field of KT did not substantially investigate KBs until 2001 [7, 21]. We reviewed the reference lists of eligible articles for additional relevant studies not identified by searches. As is typical of CIS, this was an iterative process allowing search terms to be expanded to optimize search results [26, 31].

We generated eligibility criteria based on the PICO framework (population, intervention, comparisons, and outcomes) (Additional file 3). Populations refer to MMs functioning as KBs in hospitals or other healthcare settings but did not necessarily use those labels. Because the MM literature is emergent, we included settings other than hospitals (e.g., public health department, Veteran Affairs Medical Centres). We included studies involving clinical and non-clinical administrators, managers, directors, or operational leaders if those studies met all other inclusion criteria. The intervention of interest was how MM KBs operated in practice for the creation, use and sharing of knowledge, implementation of evidence-based practice(s), or innovation implementation. Study comparisons may have evaluated one or more MM KB roles, approaches and associated barriers, enablers and impacts alone or in comparison with other types of approaches for the sharing or implementation of knowledge, evidence, evidence-based practices, or innovations. Outcomes included but were not limited to MM KB effectiveness (change in knowledge, skills, policies and/or practices, care delivery, satisfaction in role), behaviors, and outcomes. Searches were limited to English language quantitative, randomized, or pragmatic controlled trials, case studies, surveys, quasi-experimental, qualitative, or mixed methods studies and conceptual papers. Systematic reviews were not eligible, but we screened references for additional eligible primary studies. Publications in the form of editorials, abstracts, protocols, unpublished theses, conference proceedings were not eligible.

FB and ARG independently screened 50 titles and abstracts according to the eligibility criteria and compared and discussed results. Based on discrepancies, they modified the eligibility criteria and discussed how to apply them. Thereafter, FB screened all remaining titles, and discussed all uncertainties with ARG and the research team. FB retrieved all potentially eligible articles. FB and ARG independently screened a sample of 25 full-text articles, and again discussed selection discrepancies to further standardize how eligibility criteria were applied. Thereafter, FB screened all remaining full-text items.

We employed quality appraisal tools relevant to different research designs: Standards for Reporting Qualitative Research (SRQR) [32], the Good Reporting of a Mixed Methods Study (GRAMMS) tool [33], Critical Appraisal of a Questionnaire Study [34], Revised Standards for Quality Improvement Reporting Excellence (SQUIRE 2.0) tool [35], and the Critical Appraisal Checklist for Quasi-Experimental Studies [36]. FB and ARG independently assessed and compared the quality of a sample of seven studies each. Thereafter, FB assessed the quality of the remaining 24 studies.

We developed a data extraction form to extract information on study characteristics (date of publication, country, purpose, research design) and MM KB characteristics, roles, activities, enablers, barriers, and impacts. To pilot test data extraction, FB and ARG independently extracted data from the same 25 articles, then compared results and discussed how to refine data extraction. Thereafter, FB extracted data from remaining articles, which was independently checked by ARG, and then reviewed by the research team.

FB and ARG conducted an initial reading and coding of a sample of articles independently. Codes were assigned to significant elements of data within the results and conclusions sections of the eligible articles and grouped into relevant categories with shared characteristics and organized into preliminary themes. This was an iterative process that involved ongoing consultation with the research team, who provided feedback on the codes and themes.

We created a matrix of MM KB roles and activities from extant MM and KB theory [7, 18, 22, 37] and deductively mapped themes from included studies with the matrix to help inform the analysis and interpretation of our findings. As per CIS methodology, we developed an integrative grid (matrix table) where themes pertaining to MM KB roles and activities formed columns, and themes mapped to those roles/activities from individual studies formed rows [31]. The grid helped us integrate the evidence across studies and explore relationships between concepts and themes to inductively develop synthetic constructs [31, 38]. Using a constant comparative approach, we critiqued the synthetic constructs with the full sample of papers to identify conceptual gaps in the available evidence in relation to our aims, and to ensure that the constructs were grounded in the data [31, 38]. Our interpretive reflections on MM KB roles, activities, factors, and impacts led us to develop “synthetic arguments” and we used the arguments to structure our findings (attributes, roles, activities, impacts, enablers, barriers) in an organizing framework to capture our interpretation of how MMs function as KBs in healthcare organizations. We used NVivo 12 software to assist with data analysis.

The initial search yielded 9936 articles. Following removal of duplicates, 9760 titles were not eligible, and 176 items were retrieved as potentially relevant. Of those, 135 were excluded because the study design was ineligible (25), they did not examine MMs (27) or MM KBs (34), were not focused on the evaluation of an MM KB role (39), were editorials (4), or the publication was a duplicate (6). We included 41 articles for review (Fig. 1 PRISMA flow diagram). Additional file 4 includes all data extracted from included studies.

Eligible articles were published between 2003 and 2019. Three (7.3%) were conceptual and 38 (92.7%) were empirical studies. Conceptual articles discussed MM and KB theoretical constructs. Table 1 summarizes study characteristics. Studies examined the impacts of change efforts (47.3%), barriers to practice change (34.2%), and evaluation of KB interventions (18.4%). Most were qualitative (52.6%) and conducted in the United States (36.8%). Of study participants (34.2%) were MMs. In most studies, participants were nurses (63.1%) or allied health (13.2%) and based in hospitals (68.4%). Otherwise, (31.6%) were based in public health or occupational health departments, primary health care centers, Veterans Affairs Medical Centres, community care, and a senior’s care facility.

A critical analysis of the included studies revealed issues related to research design, varying from data collected from heterogeneous healthcare settings and diverse types of MMs to the type of analyses completed (e.g., qualitative, mixed methods), to the strength of conclusions drawn from a few studies’ results (e.g., correlational, or causal). Fifteen (39.5%) studies met the criteria for quality. Twenty-three (60.5%) studies had minor methodological limitations (e.g., no research paradigm identified in qualitative studies, and mixed methods studies did not describe the integration of the two methods) (Additional file 5). These methodological flaws did not warrant exclusion of any studies as they provided relevant insights regarding the emerging framework.

Seven (18.4%) studies described MM KB attributes (Table 2). Of those, 4 (10.5%) identified MM attributes, 2 (5.2%) identified KB attributes, and 1 (2.6%) identified nurse knowledge broker attributes. MM KBs were described as confident, enthusiastic, and experienced with strong research skills [41, 45]. They were also responsive and approachable, with an understanding of the complexity of an innovation and the organizational context [42‐44].

Table 3 summarizes themes pertaining to roles and activities. A total of 63 activities were grouped in the following 12 MM KB roles: (1) gather data, (2) coordinate projects, (3) monitor and evaluate the progress of a project, (4) adjust implementation to organizational context, (5) disseminate information, (6) facilitate networks, (7) bridge the evidence-to-practice gap, (8) engage stakeholders, (9) convince others of the need for, and benefit of a project, (10) coach staff, (11) provide tools and resources and (12) function as a strategic influencer. Roles did not differ among MM KBs in hospital and non-hospital settings.

Table 4 summarizes the frequency of each of the 12 MM KB roles across included studies. The two most common MM KB roles were to monitor and evaluate the progress of a project (14, 36.8%) [40, 41, 47‐51, 54, 57, 60, 63‐66] and to convince others of the need for, and benefit of a project (12, 31.6%) [46‐48, 50, 51, 55, 58, 61, 64‐67]. For example, MM KBs played an important role in monitoring the progress of projects to evaluate and reinforce practice change [41, 50]. To convince others of the need for, and benefit of a project and to promote staff buy-in, they held ongoing conversations with staff to help them understand the rationale for change, reinforce the message, and encourage staff to consistently maintain the innovations on their units [46, 48, 66]. The least common MM KB role was project coordination (4, 10.5%) [39, 47, 48, 56].

Several of the identified MM KB roles aligned with five KB roles in prior published frameworks [7, 22] and MM role theory [18, 37] (Table 5). For example, 31 (81.6%) studies described MM KB roles of gather data, project coordination, disseminate information, and adjust implementation to organizational context, which aligned with the roles and activities of a KB knowledge manager. Twenty-nine (76.3%) studies described the MM KB roles of provide tools and resources, convince others of the need for and benefit of a project, and coach staff, which aligned with the roles and activities of a KB capacity builder. We found overlap between the MM KB roles and the four hypothesized roles in MM role theory: (1) disseminate and obtain information, (2) adapt information and the innovations, (3) mediate between strategy and day to day activities, and (4) selling innovation implementation) [18, 37]. For example, we found that as capacity builders, MM KBs also mediated between strategy and day-to-day activities such as coaching staff and providing resources, and in the role of knowledge manager, MM KBs obtained, diffused, and synthesized information [18, 37].

While MM KB roles identified in included studies aligned with the five previously identified KB roles, the CIS approach we employed identified 12 distinct roles that were further characterized based on corresponding activities associated with each of the 12 roles. Therefore, while this research agrees with prior work on MM KB roles, it represents a robust framework of MM KB roles and activities by elaborating the complexity of MM KB roles and activities.

We fully described two roles compared with prior frameworks: to convince others of the need for and benefit of a project, and function as a strategic influencer. To convince others of the need for and benefit of a project (e.g., a quality improvement, best practice guideline implementation, or innovation), MM KBs used tactics such as role modelling their commitment, providing the rationale for the change, being enthusiastic about its adoption, offering positive reinforcement, and providing emotional support [47, 50, 58]. The role of strategic influencer featured in 7 (18.4%) studies [39, 48, 52, 56, 62, 65, 68]. For example, MM KBs were influential at the executive level of the hospital, advocating for innovations among less involved team members and administrators, including the hospital board, were members of organizational decision-making groups for strategic planning, and served as an authoritative contact for initiatives.

Table 6 summarizes the enablers and barriers of MM KB roles and activities, organized as individual or organizational factors. We identified four enablers at the organizational level: senior management support, availability of resources, engaged staff, and alignment to strategy. The most common was senior management support, featured in 12 (32.0%) studies. We found that senior management support enhanced the commitment of MM KBs to innovation implementation [16, 17, 19, 44, 45, 52, 61, 63, 66‐70]. For example, senior managers empowered and supported MM KBs to make decisions by ensuring that the necessary structures and resources were in place, and by conveying that the implementation was an organizational priority [66, 68]. We identified three individual-level facilitators: training and mentorship, personal attributes, and experience in the MM role. The most common facilitator was training and mentorship, featured in 8 (21.1%) studies. We found that training and mentorship with more experienced managers was important to the success of MM KBs and their projects, especially if they were new to their role [16, 17, 19, 41, 42, 48, 54, 68].

Studies reported more barriers (n = 8) than enablers (n = 7). We found four organizational barriers: a lack of resources, lack of senior management support, staff resistance, and a lack of time. The most common barriers were lack of resources in 12 (32.0%) studies and lack of time in 12 (32.0%) studies. A lack of resources (budget constraints, limited staff) made it challenging for MM KBs to move their projects forward [39, 42, 44, 47, 52, 55, 57, 64, 68‐71]. For example, inadequate funds interfered with obtaining appropriate resources and undermined the feasibility of implementing projects [47, 55]. In addition, staffing issues created difficulty in engaging staff in project work and low staffing levels limited capacity to provide desired standards of care [42, 64]. Additionally, a lack of protected time for data collection or other project work was identified as a significant barrier to implementing projects [17, 19, 39, 42, 44, 47, 52, 55, 57, 64, 68, 71]. MM KBs also lacked the time to nurture, support and adequately coach staff [39, 55].

We identified four individual-level barriers: lack of formal training, dissatisfaction with work life balance, being caught in the middle, and professional boundaries. The most common barriers were lack of formal training (8, 21.1%) and dissatisfaction with work life balance (8, 21.1%). For example, a lack of formal training resulted in MM KBs being unprepared for managerial roles and without the knowledge and skills to promote effective knowledge brokering and knowledge transfer with end users [17, 39, 41, 42, 55, 57, 69, 71]. We also found that heavy workloads and conflicting priorities left MM KBs often dissatisfied with their work life balance and hindered their ability to successfully complete projects [42, 44, 51, 52, 57, 61, 64, 71]. For example, because of multiple responsibilities and conflicting priorities, MM KBs were often pulled away to address problems or were so absorbed by administrative tasks that they had no time to complete project responsibilities [44, 64].

Eight (21.1%) studies showed that MM KBs had some impact on organizational and provider outcomes [16, 40, 43, 44, 47, 56, 62, 67]. One (2.6%) study reported that practice changes were greater when associated with higher MM leadership scores (OR 1.92 to 6.78) and when MMs worked to help create and sustain practice changes [40]. One (2.6%) study reported the impact of senior managers’ implementation of an evidence-based Hospital Elder Life Program on administrative outcomes (e.g., reduced length of stay and cost per patient), clinical outcomes (e.g., decreased episodes of delirium and reduced falls), and provider outcomes (e.g., increased knowledge and satisfaction) [67].

Two (5.3%) studies reported the impact of a Clinical Nurse Leader role on care processes at the service level in American hospitals. Benefits were evident in administrative outcomes such as RN hours per patient day (increased from 3.76 to 4.07) and in reduced surgical cancellation rates from 30 to 14%. There were also significantly improved patient outcomes in dementia care, pressure ulcer prevention, as well as ventilator-assisted pneumonia [56, 62]. One (2.6%) study reported financial savings [56].

Four (10.5%) studies reported the effect of a KB strategy on health professionals’ knowledge, skills, and practices [16, 43, 44, 47]. For example, Traynor et al. [44] found that participants who worked closely with a KB showed a statistically significant increase in knowledge and skill (average increase of 2.8 points out of a possible 36 (95% CI 2.0 to 3.6, p < 0.001) from baseline.

We sought to capture the roles, activities, enablers, barriers and impacts of MM KBs across diverse healthcare settings in an organizing framework (Fig. 2 Organizing framework of MMs who function as knowledge brokers in healthcare organizations). From our interpretation of the published evidence, the findings across studies were categorized into 12 roles and 63 associated activities to represent specific ways in which MM KBs described their roles and activities during project implementation. Influencing factors were categorized into individual and organizational enablers and barriers that influence the efforts of MM KBs in healthcare organizations. While attributes were categorized as enablers, their level of importance as enablers emerged from our synthesis in how they operated in practice. The types of outcomes that we examined also varied between changes in care practice, processes, and competencies which we constructed into provider and organizational outcomes. Our emergent insights were used to construct four synthesizing arguments from the available literature: (1) MM KBs have attributes that equip and motivate them to implement practice change and innovations in healthcare organizations, (2) MMs enact KB roles and activities in healthcare organizations, (3) enablers and barriers influence the knowledge brokering efforts of MMs in healthcare settings; and (4) MM KB efforts impact healthcare service delivery. These synthesizing arguments were used to structure the organizing framework presented in Fig. 2, which depicts how MM function as KBs in healthcare organizations and their impact on service delivery.